JP6409165B2 - Electronic device and handwritten character input program - Google Patents

Description

  The present invention relates to an electronic device and a handwritten character input program.

In recent years, an information processing apparatus (also referred to as an electronic device) including an input device such as a touch panel in which a display device such as an LCD (Liquid Crystal Display) is combined with a device that detects the coordinates of a contact position by a user operation has become widespread. For example, PC (Personal Computer)
A mobile phone, a smart phone, a notebook PC, a tablet PC, a PDA (Personal Data Assistance), a navigation device, and the like can be exemplified as an information processing apparatus including an input device such as a touch panel.

  In an information processing apparatus including a touch panel or the like, for example, an operation finger, an operation pen, or the like brought into contact with the touch panel or the like can be moved to input handwritten characters based on a time-series locus of the contact position. The handwritten character input can be made to function by, for example, starting application software (Application software, hereinafter also referred to as “application”) that is installed in the information processing apparatus and enables handwritten character input. In handwritten character input, for example, an online handwritten character recognition technique is known in which a time-series locus of a contact position is recognized as handwritten character coordinate data, and the recognized coordinate data is read as a character code.

  In the online handwritten character recognition technique, a technique for improving the character recognition rate as a character string by adding a post-processing for recognizing a word as a character string to the character recognition result of a handwritten character input via a touch panel or the like. As a technique for improving the character recognition rate, for example, a method of preparing a word dictionary when inputting handwritten characters and determining a character recognition candidate by combining a plurality of candidate characters has been proposed.

  In addition, the following patent documents exist as prior art documents in which technologies related to the technologies described in this specification are described.

JP 09-198404 A Japanese Patent Laid-Open No. 09-106442

  However, the recognition rate of characters input by handwriting tends to depend on the character input environment. For example, an input environment where the operator sits on a chair and the touch panel surface for inputting characters is stable without shaking, and the touch panel surface is unstable such as when standing while holding the information processing device with one hand There is a tendency that the character recognition rate in each input environment is different from the input environment.

When a word dictionary is used, when there are a plurality of candidate character groups for each character of a character string input by handwriting, the processing time tends to be longer due to the candidate character combination processing. For this reason, for example, in an information processing apparatus that executes recognition processing for a character string, the operating rate of a CPU (Central Processing Unit) or the like increases, and power consumption of a battery or the like tends to be accelerated. However, depending on the processing during handwriting input, the character recognition rate may increase without using a word dictionary. For example, it is a case where handwritten character input is performed in a state where there is no movement such as shaking of the casing of the information processing apparatus.

  In one aspect, an object of the present invention is to provide a technique for suppressing power consumption according to an environmental state at the time of handwriting input of an electronic device.

  The above technique can be exemplified by the configuration of the following electronic device. That is, the electronic device determines that the character recognition means for recognizing a plurality of candidate characters from the contact coordinates detected by the touch panel, the means for determining the movement of the casing including the touch panel, and the movement of the casing is a predetermined amount or more. A word dictionary is searched with a combination of a plurality of candidate characters recognized by the character recognition means, and a candidate character string based on the search result is output.

  According to said electronic device, the technique which suppresses power consumption according to the environmental state at the time of the handwriting input of an electronic device can be provided.

It is a figure which shows the hardware structural example of the information processing apparatus of this embodiment. It is a figure explaining the character recognition process in the handwritten character input when input environment is stable. It is a figure explaining the character recognition process in the handwritten character input when input environment is unstable. It is a figure explaining the processing block of the information processing apparatus of this embodiment. It is a figure explaining the screen state before operation of a confirmation button. It is a figure explaining the screen state after operation of a confirmation button. It is a figure which illustrates the structural example of coordinate table DB203. It is a figure which illustrates the coordinate data example of the handwriting stored in a coordinate data buffer. It is a figure which illustrates the example of data stored in a character candidate string buffer. It is a figure which shows the example of a screen before the replacement process of the candidate character by a character recognition post-process is performed. It is a figure which illustrates the example of data stored in the character candidate string buffer of the screen example of FIG. 6A. It is a figure explaining the production | generation of a candidate character string. It is a figure which illustrates the example of composition of a candidate character string. It is a figure explaining matching with the registration word registered into word dictionary DB202. It is a figure which illustrates the example of data of the character candidate buffer by which candidate ranking was replaced by the post-character recognition process. It is a figure which illustrates the example of a screen displayed, after candidate ranking is replaced by character recognition post-processing. It is a flowchart which illustrates the process which concerns on the character recognition of the handwritten input character of the information processing apparatus of this embodiment. It is a flowchart which illustrates the process which concerns on the character recognition of the handwritten input character of the information processing apparatus of this embodiment. It is a flowchart which illustrates the process which concerns on the character recognition of the handwritten input character of the information processing apparatus of this embodiment. It is a flowchart which shows the sensor process which concerns on an input environment as an example.

  Hereinafter, an electronic apparatus according to an embodiment will be described with reference to the drawings. The configuration of the following embodiment is an exemplification, and the electronic apparatus is not limited to the configuration of the embodiment.

Hereinafter, an electronic device will be described with reference to FIGS. 1 to 7.
<Example>
FIG. 1 illustrates a hardware configuration of an information processing apparatus (hereinafter also referred to as an electronic device) according to the present embodiment. The information processing apparatus 10 according to the present embodiment includes, for example, an information processing device including an input device such as a touch panel that combines a device surface of a display device such as an LCD (Liquid Crystal Display) with a device that detects coordinates of a contact position by a user operation. Device.

The information processing device 10 includes information processing devices such as a PC (PC: Personal Computer), a mobile phone, a smartphone, a notebook PC, a tablet PC, a PDA (Personal Data Assistance), and a navigation device. However, the information processing apparatus 10 may be a game machine or the like as long as it is an information processing apparatus having a handwritten character input function via an input device such as a touch panel.

  The information processing apparatus 10 illustrated in FIG. 1 includes a display device such as an LCD 15 a in the output unit 15. In addition, the information processing apparatus 10 includes a device that detects coordinates of a contact position with respect to a device surface of a display device such as the touch sensor 14a in the input unit 14. The touch sensor 14a is combined with a display device such as an LCD 15a, for example, to detect the coordinates of a contact position by an operator (hereinafter also referred to as a user) touch operation on the surface of the display device, and functions as a pointing device such as a touch panel To do.

  In the information processing apparatus 10 including a touch panel or the like, for example, an operation finger or an operation pen brought into contact with the touch panel or the like is moved, and handwritten character input based on a time-series locus (hereinafter also referred to as handwriting) of the contact position is possible. is there. The handwritten character input can be made to function by, for example, starting application software (hereinafter referred to as “application”) that is installed in the information processing apparatus 10 and enables handwritten character input.

  In the information processing apparatus 10 that has activated the application that enables handwritten character input, for example, the time-series locus of the contact position acquired via the touch sensor 14a is displayed on the device surface of the display device. Furthermore, the information processing apparatus 10 can recognize the locus as coordinate data of handwritten characters and read the recognized coordinate data as character codes.

  For example, the information processing apparatus 10 refers to the time-series trajectory of the contact position acquired via the touch sensor 14a and the character recognition dictionary DB 201 stored in the auxiliary storage unit 13, and corresponds to the input coordinate data. Identify candidate character groups for character codes. In the candidate character group of the character code corresponding to the input coordinate data, for example, the similarity (probability) indicating the degree of matching with the character shape data defined for the character code registered in the character recognition dictionary DB 201 Thus, ranking among candidate characters is performed.

  The information processing apparatus 10 further refers to, for example, the word dictionary DB 202 for a character string obtained by combining each candidate character group for each input handwritten character, and performs language processing at the word / phrase level. Run as a process.

For example, the information processing apparatus 10 replaces the candidate character group recognized by each handwritten character with respect to the input character string, and generates a combination as a character string. The information processing apparatus 10 refers to, for example, the word dictionary DB 202 and performs matching between the generated character string and a registered word registered in the word dictionary DB 202. For example, when there is a registered word that matches the generated character string in the word dictionary DB 202, the information processing apparatus 10 replaces the candidate character group candidate rank for each handwritten character based on the registered word. Do.

  In addition, for example, when a part of a character string matches a word that is a noun registered in the word dictionary DB 202, the information processing apparatus 10 preferentially selects a candidate word as a particle for the next character of the word. Apply and execute rules according to Japanese grammar such as matching. In the information processing apparatus 10, for example, the recognition accuracy of characters / character strings can be improved by executing, as post-processing, language processing at the word / phrase level including candidate character groups for individual handwritten character inputs.

  On the display screen such as the LCD 15a of the information processing apparatus 10, for example, for a character string input by handwriting, language processing at the word / phrase level is executed as post-processing including candidate character groups for individual handwritten characters. A character string recognized based on the result is displayed.

  In the following description, the candidate rank replacement process for each handwritten input character based on language processing at the word / phrase level is defined as “character recognition post-processing”. The information processing apparatus 10, for example, candidate character group candidates based on the candidate character group for each handwritten input character that has been character-recognized with reference to the character recognition dictionary 201, the word registered in the word dictionary DB 202, and the like. “Character recognition post-processing” for changing the order is executed.

  Moreover, the information processing apparatus 10 of this embodiment illustrated in FIG. 1 includes a sensor that detects an operating environment state of the information processing apparatus 10 related to input of handwritten characters, for example, an acceleration sensor 14b. The acceleration sensor 14b detects, for example, up / down / left / right front / rear direction shaking of the information processing apparatus 10 as a change in acceleration value. For example, the information processing apparatus 10 specifies whether the force environment of the information processing apparatus 10 related to input of handwritten characters is stable or unstable based on a change in acceleration value detected through the acceleration sensor 14b. To do.

  However, the sensor that detects the environmental state related to the input of handwritten characters of the information processing apparatus 10 is not limited to the acceleration sensor 14b. For example, a gravity sensor that detects the direction of gravity of the information processing device 10, a gyro sensor that detects rotation of the information processing device 10 in the vertical and horizontal directions, and the like may be used.

  2A and 2B exemplify explanatory diagrams of character recognition processing in handwritten character input in the information processing apparatus 10 of the present embodiment. 2A and 2B are examples of display screens displayed on the LCD 15a or the like when handwritten characters are input. For example, the information processing apparatus 10 detects an activation operation of an application that enables input of handwritten characters, and displays an input screen of handwritten characters illustrated in FIGS. 2A and 2B on the LCD 15a and the like. For example, the information processing apparatus 10 includes an input frame A1 for handwritten characters, a confirm button A2 that is an operation component for confirming an input character string, and a delete button A3 that is an operation component for deleting an input character. And displayed as an initial screen for prompting handwritten character input.

  2A and 2B are display screen examples in which three handwritten input characters can be input continuously, and the handwritten character input frame A1 region is further divided into character frames A1a, A1b, and A1c. ing. The handwritten characters input in the character frames A1a-A1c are recognized as, for example, a three-character string combining individual handwritten characters. The character string recognized from each handwritten character is displayed, for example, in the upper left display area A4 on the display screen. In each character frame, for example, three handwritten characters are sequentially input from the leftmost character frame A1a to the rightmost character frame A1c.

The display screen example illustrated in FIG. 2A is, for example, a case where a user sitting on a chair or the like is
This is an example in which handwritten character input using three kana characters such as “A”, “S”, and “TA” is performed in a stable input environment without shaking the touch panel surface in the vertical and horizontal directions. When sitting on a chair or the like and performing handwriting input, the touch panel surface on which characters are input is, for example, held in a certain inclination.

  In a state where the touch panel surface is stable, for example, the user can move an operation finger or the like brought into contact with the touch panel surface without being affected by fluctuations in the up / down / left / right / front / rear directions. For this reason, the movement locus on the time series of the contact position detected in the character frame A1 displayed on the LCD 15a or the like becomes a stable character shape according to the input character intended by the user. As illustrated in FIG. 2A, for example, “A”, which is a three-screen kana character in the character frame A1a, “Shi”, which is a one-screen kana character in the character frame A1b, and the character frame A1c, The 4-character kana “ta” is input in a stable character form.

  For example, the information processing apparatus 10 sets the character frames A1a to A1c based on the time series of the coordinate data of the movement trajectory moved until the touched operation finger or the like is released in each character frame region. Performs character recognition for each handwritten character entered. As a result of the character recognition, the character code corresponding to the handwritten character input in the character frame A1a-A1c is specified. Note that the character codes for the handwritten characters input in the character frames A1a-A1c are identified by, for example, a plurality of character codes according to the similarity indicating the degree of matching with the characters registered in the character recognition dictionary DB 201. Recognized as a character group. The similarity is, for example, an index representing the probability of character recognition of a candidate character for a handwritten character input in a character frame.

  In the example of FIG. 2A, the correct answer “A” is recognized as the first candidate for the handwritten character input in the character frame A1a. In addition, “o”, which is a 3-character kana character, is recognized as a second candidate, and “su”, which is a 2-screen kana character, is recognized as a third candidate. Similarly, “shi”, which is the correct answer, is recognized as the first candidate for the handwritten character input in the character frame A1b. In addition, “ku” which is a kana character of one stroke is recognized as a second candidate, and “re” which is a katakana character of one stroke is recognized as a third candidate. For the handwritten character input in the character frame A1c, the correct answer “TA” is recognized as the first candidate. In addition, “Ni”, which is a kana character of three strokes, is recognized as a second candidate, and “n”, which is a kana character of one stroke, is recognized as a third candidate.

  2A has display areas A5-A7 for displaying candidate character groups recognized for each character frame below the display area of each character frame. For example, the first candidate character recognized in each character frame is displayed in the display area A5. Similarly, for example, the second candidate character recognized in each character frame is displayed in the display area A6, and the third candidate character recognized in each character frame is displayed in the display area A7. In the display screen example of FIG. 2A, a character string including a candidate character recognized as the first candidate for the handwritten character input in the character frame A1a-A1c is displayed in the display area A4 in the upper left corner. Yes.

In the display area A4 of FIG. 2A, the character string “ Ashita ” including the candidate character recognized as the first candidate is displayed with an underline in the handwritten characters input in the character frames A1a to A1c. Note that the underline added to the character string displayed in the display area A4 represents an undetermined character recognition state, and is deleted, for example, in response to a user operation on the confirmation button A2. In the display area A4 after the confirmation operation, for example, a character string “Ashita” in a state where the underline is deleted is displayed.

As illustrated in FIG. 2A, in an input environment in which stable handwriting input can be performed on the touch panel surface, the shape of the handwritten character input in accordance with the user's intention is stable. For this reason, in the character candidate group based on the time-series coordinate data as the movement locus of the contact position, the correct character code tends to appear as the first candidate. Moreover, the similarity of each candidate character tends to be relatively low for other candidate characters other than the first candidate. In the information processing apparatus 10 in a stable input environment, there is a tendency that a character code that is correct with respect to an input handwritten character is included in a higher candidate as a character recognition result.

  Next, character recognition processing in an unstable input environment in which the touch panel surface fluctuates in the up / down / left / right / front / rear directions will be described with reference to the drawing illustrated in FIG. 2B. An example of the display screen illustrated in FIG. 2B is, for example, a handwritten character using three kana characters such as “a”, “shi”, “ta”, etc. while the user holding the information processing apparatus 10 is standing. This is an example when input is performed.

  In the case where handwritten character input is performed in a standing position, for example, the information processing apparatus 10 is held by the other hand different from the operating hand performing character input. For this reason, the touch panel surface of the information processing apparatus 10 held with one hand becomes an unstable input environment such as shaking up and down, left and right, and back and forth, or the tilt of the touch panel surface being unstable.

  In a state where the touch panel surface is unstable, the user moves, for example, an operation finger or the like that is in contact with the touch panel surface while being affected by up / down, left / right, front / rear direction shaking, unstable tilt, or the like. For this reason, the movement trajectory in the time series of the contact position detected in the character frame A1 displayed on the LCD 15a or the like is distorted with respect to the input character intended by the user. Further, for example, as illustrated in the character frame A1a of FIG. 2B, there is a possibility that an erroneous input A8 caused by an erroneous contact may occur.

  In the information processing apparatus 10 in which the character shape disorder or incorrect input has occurred, for example, in the time series of the coordinate data of the movement trajectory moved until the touched operating finger or the like is released in each character frame area. Based on this, the character recognition of the handwritten character input into each character frame is performed. In the information processing apparatus 10, for example, character recognition of handwritten characters input for each character frame is performed based on the time series of the coordinate data of the movement trajectory in a state where a character shape disorder or erroneous input has occurred.

  As a result of the character recognition, as illustrated in FIG. 2B, for the input of the kana character “a” in which the erroneous input A8 of the character frame A1a has occurred, the three-character kana character “o” is recognized as the first candidate. The Rukoto. The correct answer “a” is recognized as the second candidate, and “ra”, which is a two-stroke kana character, is recognized as the third candidate. In the character frame A1b, the correct answer “shi” is recognized as the first candidate for the input of the kana character “shi” whose shape is disordered. Further, “Re”, which is a one-stroke katakana character, is recognized as a second candidate, and “ku”, which is a one-stroke kana character, is recognized as a third candidate. In the character frame A1c, “n”, which is a single-character kana character, is recognized as the first candidate in response to the input of the kana character “TA” whose character shape is disordered. Note that “Ni”, which is a kana character of two strokes, is recognized as the second candidate, and “T”, which is the correct answer, is recognized as the third candidate.

  In the information processing apparatus 10 in which the character shape disorder or incorrect input has occurred, for example, “Oshin”, which is a character string including the character code recognized as the first candidate, is displayed in the display area A4 in the upper left corner. Become.

  As illustrated in FIG. 2B, in the case where the touch panel surface for inputting handwritten characters fluctuates in the up / down / left / right / front / rear direction, or in an unstable input environment such as the tilt of the touch panel surface being unstable, the user's intention is satisfied. The shape of the handwritten characters entered in the above becomes unstable. Moreover, there is a possibility of including an erroneous input caused by an erroneous contact of an operation finger or the like. For this reason, in the candidate character group representing the character code recognized based on the time-series coordinate data of the movement locus, the correct character code tends not to appear in the first candidate. In addition, the similarity of each character appearing as a candidate character to the correct candidate tends to be gathered with a low similarity overall.

In the information processing apparatus 10, for example, after the candidate character group for the handwritten input character illustrated in FIGS. 2A and 2B is specified, the word dictionary DB 202 is referred to, and the character recognition post-processing is executed. In the post-character recognition processing, for example, candidate character combinations are performed after the candidate character group candidate ranks of individual handwritten characters recognized in the character frames A1a to A1c are replaced, and the word dictionary DB 202 is used as a character string. The matching process with the registered word registered in the above is executed.

  For example, in the example of FIG. 2A, there are three candidate character groups recognized in the character frame A1a: the first candidate “A”, the second candidate “O”, and the third candidate “SU”. There are three candidate character groups recognized in A1b: the first candidate “shi”, the second candidate “ku”, and the third candidate “le”. There are three candidate character groups recognized in the character frame A1c: the first candidate “ta”, the second candidate “ni”, and the third candidate “n”.

  The information processing apparatus 10 generates, for example, three character strings of “3 types × 3 types × 3 types = 27 types” by exchanging the candidate rankings of the three candidate character groups recognized in each character frame and combining them. To do. Then, for example, the information processing apparatus 10 executes matching for matching each of the generated 27 character strings with registered words registered in the word dictionary DB 202.

  Similarly, in the example of FIG. 2B, the information processing apparatus 10 switches the candidate ranks of the three candidate character groups recognized in each of the character frames A1a, A1b, and A1c, for example, and combines 27 candidate three-character combinations. A string is generated. For example, the three candidate character groups {a, o, ra} in the character frame A1a, the three candidate character groups in the character frame A1b, and the three candidate character groups in the character frame A1c {n ,,} Are combined to generate 27 different three-character strings. Then, for example, the information processing apparatus 10 performs matching between each of the generated 27 character strings and a registered word registered in the word dictionary DB 202.

  As a result of the character recognition post-processing, the information processing apparatus 10 recognizes a character string {Ashita} using three kana characters that are correct, based on language processing at the word / phrase level, for example.

  Here, in the case of FIG. 2A, for example, a handwritten character included in the correct character string is recognized as the first candidate character for the correct three character character string {Ashita}. For this reason, in the post-character recognition processing executed by the information processing apparatus 10, for example, the correct combination of the first candidate characters can be obtained without performing the processing for replacing candidate ranks of candidate character groups recognized by individual handwritten characters. It is possible to recognize character strings.

  On the other hand, in the case of FIG. 2B, for example, the correct answer is obtained by combining the candidate characters recognized as the second candidate character, the first candidate character, and the third candidate character in the order of the character frames A1a, A1b, and A1c. A 3-character string {Ashita} can be recognized. In the case of FIG. 2B, for example, the final character recognition for a candidate character is performed by executing a character recognition post-processing that replaces the candidate character group candidate ranks for each recognized handwritten character and combines them as a character string. It can be said that the rate can be increased.

However, in the post-character recognition process, the candidate ranks of candidate character groups recognized by individual handwritten characters are replaced and combined as character strings, so that the processing time until matching with the correct character string is long. Tend to be. For this reason, when performing post-character recognition processing, the processing load and operation rate of a CPU (Central Processing Unit) and the like related to post-character recognition processing increase, and power consumption of the battery and the like tends to be accelerated.

In order to shorten the processing time related to the post-character recognition processing, for example, it is conceivable to reduce the number of candidate character groups recognized by individual handwritten characters. For example, in the example of FIG. 2B, the processing target of the character recognition post-processing can be exemplified as the first candidate and the second candidate. However, as described above, it is the third candidate character that appears as the third character of the correct character string {Ashita}. For this reason, even if the number of candidates to be processed in the character recognition post-processing is reduced, in the example of FIG. 2B, it becomes difficult to recognize the correct character string {Ashita}, resulting in a decrease in the character recognition rate. Become. This is because, as illustrated in FIG. 2B, in a state where the input environment for handwritten character input is unstable, the similarities to the correct answers of characters recognized as candidate characters tend to be gathered with a relatively low similarity.

  In the information processing apparatus 10 according to the present embodiment, for example, the state (input environment) of the information processing apparatus 10 related to handwritten character input is in a stable state or an unstable state based on a change in acceleration value detected from the acceleration sensor 14b. Determine if there is. When the state of the information processing apparatus 10 related to handwritten character input is in an unstable state, the information processing apparatus 10 replaces the candidate ranks of candidate character groups recognized from individual handwritten characters, for example, Executes character recognition post-processing to be combined as a column. As a result, in the information processing apparatus 10 of the present embodiment, it is possible to suppress power consumption related to post-character recognition processing. In the information processing apparatus 10 according to the present embodiment, for example, a technique for suppressing power consumption can be realized according to an environmental state at the time of handwriting input of an electronic device.

〔Device configuration〕
Returning to FIG. 1, the information processing apparatus 10 of this embodiment includes a CPU (Central Processing Unit) 11, a main storage unit 12, an auxiliary storage unit 13, and an input unit 1 that are connected to each other via a connection bus B <b> 1.
4, an output unit 15, and a communication unit 16. The main storage unit 12 and the auxiliary storage unit 13 are recording media that can be read by the information processing apparatus 10.

  In the information processing apparatus 10, the CPU 11 develops a program stored in the auxiliary storage unit 13 so as to be executable in the work area of the main storage unit 12, and controls peripheral devices through the execution of the program. Thereby, the information processing apparatus 10 can implement a function that matches the predetermined purpose described above.

  In the information processing apparatus 10 illustrated in FIG. 1, the CPU 11 is a central processing unit that controls the entire information processing apparatus 10. The CPU 11 performs processing according to a program stored in the auxiliary storage unit 13. The main storage unit 12 is a storage medium in which the CPU 11 caches programs and data and expands a work area. The main storage unit 12 includes, for example, a RAM (Random Access Memory) and a ROM (Read Only Memory).

  The auxiliary storage unit 13 stores various programs and various data in a recording medium in a readable and writable manner. The auxiliary storage unit 13 is also called an external storage device. The auxiliary storage unit 13 stores an OS, various programs, various tables, and the like. For example, the OS provides the application with an interface to a resource to be managed for the installed application. An application installed in the terminal device 10 realizes an application function by using an interface to a resource provided by the OS. The OS includes a communication interface program that exchanges data with an external device or the like connected via the communication unit 16. Examples of the external device include other information processing devices such as a PC and a server, an external storage device, and the like on a network (not shown).

The auxiliary storage unit 13 is, for example, an EPROM (Erasable Programmable ROM), a solid state drive device, a hard disk drive (HDD, Hard Disk Drive) device, or the like. As the auxiliary storage unit 13, for example, a CD drive device, a DVD drive device, a BD drive device, or the like can be presented. Examples of the recording medium include a silicon disk including a nonvolatile semiconductor memory (flash memory), a hard disk, a CD, a DVD, a BD, a USB (Universal Serial Bus) memory, and a memory card.

The input unit 14 receives an operation instruction or the like from a user or the like. The input unit 14 is an input device such as a camera, an input button, a touch sensor 14a, an acceleration sensor 14b, a pointing device, or a microphone. The input unit 14 may include a keyboard, a wireless remote controller, and the like. The pointing device includes, for example, a touch panel, a mouse, a trackball, a joystick, etc. in which the touch sensor 14a and the LCD 15a are combined. Information input from the input unit 14 is notified to the CPU 11 via the connection bus B1. For example, coordinate information detected by the pointing device, acceleration information detected by the acceleration sensor 14b, and the like are notified to the CPU 11 via the connection bus B1.

The output unit 15 outputs data processed by the CPU 11 and data stored in the main storage unit 12. The output unit 15 includes display devices such as an LCD 15a, a CRT (Cathode Ray Tube) display, a PDP (Plasma Display Panel), an EL (Electroluminescence) panel, and an organic EL panel. The output unit 15 includes output devices such as a printer and a speaker.

  The communication unit 16 is an interface with, for example, a network to which the information processing apparatus 10 is connected. The network includes, for example, a purchase network such as the Internet, a wireless network such as a mobile phone network, and a LAN (Local Area Network).

  In the information processing apparatus 10, the CPU 11 reads out the OS, various programs, and various data stored in the auxiliary storage unit 13 to the main storage unit 12 and executes them. Realize the means. The information processing apparatus 10 implements the initial setting unit 101, the coordinate input unit 102, the character recognition unit 103, and the display control unit 104 illustrated in FIG. 1 by executing the target program. Note that the coordinate input unit 102 includes a timer 102a. The timer 102 a is a timer function provided in the information processing apparatus 10. The character recognition unit 103 includes an acceleration sensor processing unit 103a. However, any of the processing means illustrated in FIG. 1 or a part of them may be operated by a hardware circuit.

  The information processing apparatus 10 includes, for example, a character recognition dictionary 201, a word dictionary 202, a coordinate table DB 203, and a screen image DB 204 as storage destinations of data to be referred to or managed by each of the above processing units. Prepare for.

  Any one of the processing units illustrated in FIG. 1 may be included in another information processing apparatus or the like. For example, an information processing device including the character recognition unit 103 and an information processing device including the initial setting unit 101, the coordinate input unit 102, and the display control unit 104 may be connected via a network and function as the information processing device 10. . The information processing apparatus 10 can be realized as a cloud that is a group of computers on a network, for example.

[Process block configuration]
(Initial setting part)
FIG. 3 illustrates a processing block diagram illustrating the processing of the information processing apparatus 10 according to the present embodiment. In the processing block diagram illustrated in FIG. 3, for example, the initial setting unit 101 refers to the screen image DB 204 in response to an activation operation of an application that enables handwritten character input, and recognizes input handwritten characters. Is displayed on a display screen such as the LCD 15a. Further, the initial setting unit 101 performs initialization processing such as various buffer areas for processing provided in a predetermined area of the main storage unit 12 for character recognition processing, for example.

The initial screen for recognizing handwritten character input is, for example, as illustrated in FIGS. 2A and 2B, an input frame A1 for handwritten characters, a confirm button A2 which is an operation component for confirming an input character string, It includes a delete button A3 that is an operation component for deleting an input character. The area of the input frame A1 may be further divided into a plurality of character frame areas. 2A and 2B are examples in which the area of the input frame A1 is divided into three adjacent areas of character frames A1a, A1b, and A1c.

  In addition, in the initial screen, as illustrated in FIG. 2A and the like, a display area that displays candidate characters recognized for the handwritten characters input in each character frame in a lower area of the display area of each divided character frame. A5-A7 may be provided. Character recognition results for handwriting of handwritten characters input in each character frame can be displayed as a plurality of candidate characters.

  In addition, on the initial screen, as illustrated in FIG. 2A and the like, there is a display area A4 for displaying a character string including a handwritten character recognized as the first candidate for the handwritten character input in the character frame A1a-A1c. It may be provided. The character string displayed in the display area A4 may be subjected to character modification indicating that the character string is in an undetermined character recognition state. Examples of character modification indicating an unconfirmed character recognition state include underlining, bold modification, oblique character modification, and hatching. FIG. 2A and the like are examples in which underlining is applied as character modification indicating that the character recognition state is undetermined.

  In addition, the character modification with respect to the character string displayed on the display area may be deleted, for example, triggered by a user operation on the confirmation button A2. Alternatively, the display of the handwritten characters input in the character frames A1a-A1c and the candidate characters displayed in the display area A5-A7 may be erased with a user operation on the confirmation button A2. 4A and 4B illustrate transition examples of display screens before and after a user operation on the confirmation button A2.

FIG. 4A is an example of a display screen before a user operation for the confirmation button A2, and FIG. 4B is an example of a display screen after a user operation for the confirmation button A2. As illustrated in FIG. 4A, before the confirmation button A2 is operated, the display area A4 includes a character string “ Ashita ” that is a combination of characters recognized as the first candidate for the handwritten character input in each character frame. , Underlined and modified. In addition, handwriting of the input handwritten character is displayed in each character frame. In the display area A5-A7, which is the lower area of the display area of each character frame, the first to third candidate characters recognized from the handwriting of the handwritten character input to each character frame are displayed. .

On the other hand, in the display area A4 in FIG. 4B, the underline modifier added to the character string “ Ashita ” displayed in the display area A4 in FIG. 4A is deleted, and the character string “Ashita” fixed by the user operation is displayed. Yes. Moreover, the input handwriting of the handwritten character displayed in each character frame is deleted. Similarly, the first to third candidate characters displayed in the display area A5-A7, which is the lower area of each character frame, are deleted.

  In addition, in the user operation of the delete button A3 that is an operation component for deleting the input character, for example, the handwriting of the character frame that has been input most recently may be deleted, and the character after the input to all the character frames is completed. The handwriting input as a column may be deleted. Further, for example, a character frame after handwriting input may be selected and the input handwriting of the selected character frame may be deleted.

(Coordinate input part)
Returning to the processing block diagram illustrated in FIG. 3, the coordinate input unit 102 manages the display coordinates of the initial screen for recognizing handwritten character input, for example. For example, the coordinate input unit 102 refers to the coordinate table DB 203 in response to an activation operation of an application that enables handwritten character input, and display positions of various display components, input frames, various display areas, and the like displayed on the initial screen. The coordinate data indicating is acquired. For example, the coordinate input unit 102 temporarily stores the acquired coordinate data related to the initial screen in a predetermined area of the main storage unit 12.

  FIG. 5A illustrates a configuration example of the coordinate table DB 203. FIG. 5A is a configuration example of a coordinate table for the display screen illustrated in FIG. 2A, for example.

  The configuration example of the coordinate table DB 203 illustrated in FIG. 5A includes columns such as “part”, “shape”, “end point 1”, “end point 2”, “end point 3”, and “end point 4”. The “component” column stores names of display components and the like displayed on the initial screen. The “shape” column stores the shape, attributes, and the like of a display area such as a display component on the initial screen. In the “end point 1”, “end point 2”, “end point 3”, and “end point 4” columns, coordinate information corresponding to the shape and attributes of the display area stored in the “shape” column is stored. For example, when the shape of the display area stored in the “shape” column is “rectangular”, the coordinate information of the upper left corner of the display area is stored in the “endpoint 1” column, and the “endpoint 2” column is stored in the “endpoint 2” column. The coordinate information of the lower right corner of the display area is stored.

  Note that the coordinate information stored in the “end point 1” column or the like has, for example, the upper left corner of the display device such as the LCD 15a as the origin, the horizontal direction of the display device as the X axis, and the vertical direction as the Y axis (X, Y) as two-dimensional coordinates.

  In the configuration example of FIG. 5A, the coordinate table DB 203 has a record for each display component displayed on the initial screen. In the configuration example of FIG. 5A, for example, “rectangle” representing the shape of the display area is stored in the “shape” column of the record in which the “confirmation button” is stored in the “component” column. Also, for example, “(5, 20)” representing the coordinates of the upper left corner of the display area is stored in the “end point 1” column of the record in which the “confirmation button” is stored in the “component” column. The “2” column stores “(35, 30)” representing the coordinates of the lower right corner of the display area.

  Further, for example, “rectangle” representing the shape of the display area is stored in the “shape” column of the record in which the “delete button” is stored in the “component” column. Also, “(5, 35)” representing the coordinates of the upper left corner of the display area is stored in the “End Point 1” column of the record, and the coordinates of the lower right corner of the display area are stored in the “End Point 2” column. “(35, 45)” is stored. In the initial screen, it can be seen that the “confirm button” and the “delete button” are displayed with the left and right display positions in the left and right directions in common.

  In the configuration example of FIG. 5A, “input box 1” stored in the “component” column represents, for example, the character box A1a illustrated in FIG. 2A and the like. Similarly, “input box 2”, “input box” 3 ″ represents character frames A1b and A1c illustrated in FIG. 2A and the like, respectively. Similarly, the “input frame 1 first candidate” stored in the “component” column represents, for example, the first candidate of the handwritten character input in the character frame A1a illustrated in FIG. 2A or the like.

  Returning to the processing block diagram illustrated in FIG. 3, the coordinate input unit 102 manages, for example, handwriting coordinate information about handwritten characters input to the character frames A1a to A1c. The coordinate input unit 102 acquires, for example, the contact position of the operation finger or the like with respect to the area in each character frame as the handwriting start position via the touch sensor 14a. Then, the coordinate input unit 102, for example, the time-series coordinate data of the contact position related to the movement trajectory until the operation finger or the like brought into contact with the area in each character frame is detached from the display device surface via the touch sensor 14a. To get.

  For example, the coordinate input unit 102 stores time-series coordinate data of the contact position acquired via the touch sensor 14a in a coordinate data buffer provided in a predetermined area of the main storage unit 12 in time-series order. For example, the coordinate input unit 102 stores time-series coordinate data of the contact position acquired via the touch sensor 14a in the coordinate data buffer in association with each character frame.

FIG. 5B illustrates an example of handwriting coordinate data stored in the coordinate data buffer. The coordinate data stored in the coordinate data buffer is, for example, a two-dimensional (X, Y) with the upper left corner of the display device such as the LCD 15a as the origin, the horizontal direction of the display device as the X axis, and the vertical direction as the Y axis. Expressed as coordinates.

  In the example of FIG. 5B, the two-dimensional coordinates indicated by d1-d4 and d6-d7 are coordinate data sampled at a constant periodic interval of 20 ms. Note that the coordinate data d5 indicated by the two-dimensional coordinates “(0, 0)” is a delimiter inserted when the operating finger or the like is detached from the display device surface, for example.

  For example, the coordinate input unit 102 specifies that coordinate data sampled at regular intervals is not continuously acquired from the touch sensor 14a. This is because, when the coordinate data sampled at a constant cycle interval is not continuously acquired, for example, it can be determined that the operating finger or the like brought into contact with the display device is detached.

  Then, for example, as illustrated in d5 of FIG. 5B, the coordinate input unit 102 adds the delimiter indicated by the two-dimensional coordinates “(0, 0)” to the coordinate data immediately before leaving, and stores it in the coordinate data buffer. . For example, a stroke of a character input by handwriting can be identified by the delimiter inserted in the coordinate data buffer.

  For example, in the example of FIG. 5B, the coordinate data string represented by d1-d4 represents the first stroke of the handwritten character, d5 as a delimiter is inserted, and the coordinate data string after d6 is handwritten. It can be seen that it represents the handwriting of the second and subsequent strokes. In the coordinate data representing the handwriting of the first stroke, for example, the two-dimensional coordinate “(5, 60)” of d1 represents the start position of the handwriting of the first stroke, and the two-dimensional coordinate “10, 58” of d4 is represented. It can be seen that it represents the end position of the handwriting of the first stroke.

  Returning to the processing block diagram illustrated in FIG. 3, the coordinate input unit 102 makes contact with the display device surface again when the operation finger or the like whose contact position is detected in each character frame displayed on the display screen such as the LCD 15 a is released. Manage the withdrawal period until the position is detected. For example, the information processing apparatus 10 causes the timer 102a to function, and measures the separation period until the operation finger or the like whose contact position is detected in each character frame is detached and the contact position on the display device surface is detected again. . The coordinate input unit 102 manages, for example, whether the withdrawal period measured by the timer 102a is equal to or greater than a predetermined threshold. Here, the predetermined threshold value for the withdrawal period is obtained by, for example, displaying the initial screen illustrated in FIG. 2A on the display screen of the LCD 15a or the like, and experimentally inputting a character string including a predetermined kana character or the like. The withdrawal period can be measured, and the measured withdrawal period can be used as a threshold value.

  The coordinate input unit 102 can identify the completion of handwritten character input for the character frame in which the contact position is detected by managing the separation period of the operating finger or the like that is in contact with each character frame. For example, when the measured leaving period is equal to or greater than a predetermined threshold, the coordinate input unit 102 sends the coordinate data stored in the coordinate data buffer to the character recognition unit 103 for the character frame in which handwritten character input is performed. hand over.

(Character recognition part)
For example, the acceleration sensor processing unit 103a of the character recognition unit 103 identifies up / down, left / right, front / rear direction shaking of the information processing apparatus 10 based on a change in acceleration value detected by the acceleration sensor 14b. For example, the acceleration sensor processing unit 103a acquires the acceleration values in the up / down / left / right / front / rear directions detected by the acceleration sensor 14b at regular intervals. And the acceleration sensor process part 103a compares with the predetermined threshold value about the acquired acceleration value of each direction, for example, and the state of the information processing apparatus 10 which concerns on handwriting input is a stable state or an unstable state Determine whether. The result of the determination is accumulated as a flag value represented by a binary state of “0” and “1”, for example.

  Here, for example, the predetermined threshold value is obtained from acceleration values acquired in each state by performing a handwritten character input operation of the information processing apparatus 10 in an experimentally seated state or in a standing state. be able to.

  For example, when the acquired acceleration value in each direction is within a predetermined threshold, the acceleration sensor processing unit 103a sets a flag value “0” indicating that the device is in a predetermined area in the predetermined area of the main storage unit 12. Temporarily store. In addition, the acceleration sensor processing unit 103a sets a flag value “1” indicating that the device is in an unstable state to a predetermined value in the main storage unit 12 when the acquired acceleration value in each direction is equal to or greater than a predetermined threshold value. Store temporarily in the area. In the following description, the flag value indicating the state of the information processing apparatus 10 is also referred to as “terminal state value”.

  For example, the character recognition unit 103 performs character recognition based on the coordinate data representing the handwriting of characters handwritten in each character frame delivered from the coordinate input unit 102, and specifies the corresponding candidate character group To do. Character recognition for characters input by handwriting is performed, for example, by pattern matching with registered characters registered in the character recognition dictionary DB 201 in advance. For example, the character recognition unit 103 refers to the character recognition dictionary DB 201 and collates coordinate data for each character frame delivered from the coordinate input unit 102 with registered characters. As a result of the collation, the character recognition unit 103 identifies a candidate character group according to, for example, the degree of similarity with the registered character. For example, a candidate rank corresponding to the degree of similarity is given to the identified candidate character group.

  The character recognition unit 103 stores, for example, a character candidate group identified with reference to the character recognition dictionary DB 201 as a character recognition result in a character candidate string buffer provided in a predetermined area of the main storage unit 12. For example, the character recognizing unit 103 stores the candidate characters in the character candidate string buffer in association with the candidate characters in the order of candidates assigned according to the similarity. In addition, the character recognition part 103 delivers the character candidate group recognized for every character frame to the display control part 104 with the candidate order | rank provided to the character candidate, for example. The character candidate group for each character frame delivered to the display control unit 104 is displayed as a candidate character in a predetermined display area of the display screen such as the LCD 15a according to the candidate rank.

  FIG. 5C illustrates an example of data stored in the character candidate string buffer. FIG. 5C shows an example in which three candidate characters are displayed as a character recognition result for a character input by handwriting, as exemplified in FIG. 2A and the like. The data example in FIG. 5C is the character recognition example described in FIG. 2A.

  The configuration example of the data stored in the character candidate string buffer in FIG. 5C has a record for each candidate character arranged in candidate order. Each record of the configuration example has columns of “input frame 1”, “input frame 2”, and “input frame 3”. The “input frame 1”-“input frame 3” columns have sub-columns of “candidate character” and “similarity”, respectively. Note that “input frame 1”-“input frame 3” respectively correspond to the character frames A1a-A1c illustrated in FIG. 2A and the like. The “input frame 1” column stores character recognition results for handwritten characters input in the character frame A1a. Similarly, in the “input frame 2” column, the character recognition result for the handwritten character input in the character frame A1b, and in the “input frame 3” column, the character recognition result for the handwritten character input in the character frame A1c. Stored.

The “candidate characters” column stores candidate characters recognized for the handwritten characters input in the respective character frames. The “similarity” column stores an index value indicating the probability of the candidate character recognized for the handwritten character input in each character frame. For example, the degree of matching with a registered character registered in the character recognition dictionary DB 201 that is a candidate character is stored. In addition, as an index value indicating the likelihood of a candidate character, for example, a value indicating a difference between a registered character registered in the character recognition dictionary DB 201 that has become a candidate character and coordinate data acquired in the character frame is stored. You may do that. When the value indicating the difference between the character candidate and the coordinate data acquired in the character frame is adopted as the index value indicating the probability of the candidate character, for example, the smaller the stored value (the lower) the candidate character The similarity with the handwriting of the input handwritten character becomes high.

  In the configuration example of FIG. 5C, “A” displayed as the first candidate is stored in the “candidate character” sub-column of the “input frame 1” column of the first candidate record, as illustrated in FIG. 2A. Yes. In the “similarity” sub-column of the “input box 1” column of the record, “0” that is a value indicating a difference between the candidate character and the coordinate data acquired in the character box A1a is stored.

  Similarly, “shi” displayed as the first candidate is stored in the “candidate character” sub-column of the “input frame 2” column of the first candidate record, as illustrated in FIG. 2A. In the “similarity” sub-column of the “input box 2” column of the record, “1” that is a value indicating a difference between the candidate character and the coordinate data acquired in the character box A1b is stored.

  Further, in the “candidate character” sub-column of the “input frame 3” column of the first candidate record, “TA” displayed as the first candidate is stored as illustrated in FIG. 2A. In the “similarity” sub-column of the “input box 3” column of the record, “1” that is a value indicating a difference between the candidate character and the coordinate data acquired in the character box A1c is stored.

  As described with reference to FIG. 2A, when handwritten character input is performed while sitting on a chair or the like, for example, the touch panel surface or the like is in a stable state without being shaken in the up / down / left / right / front / back directions. It can be seen that each candidate character has a high similarity.

  Further, in the configuration example of FIG. 5C, “su” displayed as the third candidate is stored in the “candidate character” sub column of the “input frame 1” column of the third candidate record, as illustrated in FIG. 2A. Has been. In the “similarity” subcolumn of the “input frame 1” column of the record, “8” that is a value indicating a difference between the candidate character and the coordinate data acquired in the character frame A1a is stored.

  Similarly, “re” displayed as the third candidate is stored in the “candidate character” sub-column of the “input frame 2” column of the third candidate record, as illustrated in FIG. 2A. In the “similarity” sub column of the “input box 2” column of the record, “3” that is a value indicating a difference between the candidate character and the coordinate data acquired in the character box A1b is stored.

  In addition, “n” displayed as the third candidate is stored in the “candidate character” sub-column of the “input frame 3” column of the third candidate record, as illustrated in FIG. 2A. In the “similarity” sub-column of the “input box 3” column of the record, “11” which is a value indicating a difference between the candidate character and the coordinate data acquired in the character box A1c is stored.

  As described with reference to FIG. 2A, when handwritten character input is performed in a stable input environment, it can be seen that the lower the candidate ranking, the lower the degree of similarity recognized as a candidate character. Further, it can be seen that the difference in similarity between the character candidate recognized as the first candidate and the character candidate recognized as the third candidate is relatively large.

However, as described with reference to FIG. 2B, handwritten character input is executed even in an unstable input environment in which the touch panel surface swings up, down, left, and right, back and forth, for example. For example, handwritten characters input under an unstable input environment tend to be distorted compared to handwritten characters input under a stable environment, even for the same user. In addition, there is a tendency for erroneous input due to erroneous contact with an operation finger or the like.

  In order to improve the character recognition rate for handwritten characters input in an unstable input environment, the character recognition unit 103 uses, for example, character strings of individual handwritten characters input in each character frame, Execute character recognition post-processing to switch the candidate order. In the character recognition post-processing, the character recognition unit 103 performs a matching process with registered words registered in the word dictionary DB 202, for example, by replacing the candidate ranks of candidate character groups recognized for individual handwritten characters. For example, when there is a character string that matches a registered word registered in the word dictionary DB 202, the character recognition unit 103 outputs the replaced candidate character to the display control unit 104 as a first candidate as a character recognition result. To do. As a result of the post-character recognition process, for example, on the display screen such as the LCD 15a, the candidate character that has been subjected to the replacement process based on the character string that matches the registered word registered in the word dictionary DB 202 is displayed as the first candidate. Is done.

  Returning to the processing block diagram illustrated in FIG. 3, the character recognition unit 103 refers to, for example, the word dictionary DB 202 and performs character recognition post-processing using a candidate character group for each handwritten character input to each character frame. Run. The post-character recognition process is executed in accordance with, for example, a flag value stored in a predetermined area of the main storage unit 12 as a terminal state value. Since the character recognition post-processing uses a candidate character group recognized by each handwritten character input in each character frame, for example, it is executed for the handwritten character input as the first character of the character string in the character frame A1a. Not.

  Hereinafter, post-character recognition processing executed by the character recognition unit 103 will be described with reference to the drawings illustrated in FIGS. 6A to 6G. FIG. 6A is an example of a screen before the candidate character replacement process by the character recognition post-process is executed. In the screen example of FIG. 6A, for example, on the initial screen displayed on the LCD 15a or the like, the kana characters “a” and “me” are input by handwriting in each of the character frames A1a and A1b, and each character is verified by collation with the character recognition dictionary It is an example of a screen in the state where the candidate character group about is displayed.

  In the screen example of FIG. 6A, “O” is displayed as a first candidate character in the display area A5 for the kana character “A” input by handwriting in the character frame A1a. Similarly, the candidate character “a” of the second candidate is displayed in the display area A6, and the candidate character “chi” of the third candidate is displayed in the display area A7.

In addition, “M” is displayed as a first candidate character in the display area A5 for the kana character “M” input by handwriting in the character frame A1b. Similarly, the second candidate character “N” is displayed in the display area A6, and the third candidate character “A” is displayed in the display area A7. In addition, in the display area A4 of FIG. 6A, “ Ome ” for the character string displayed as the first candidate is displayed in an underlined state indicating an unconfirmed state.

  FIG. 6B illustrates an example of data stored in the character candidate string buffer for the screen example of FIG. 6A. In the character candidate string buffer illustrated in FIG. 6B, as described with reference to FIG. 5C, the records of candidate characters recognized for the characters input by handwriting in the character frames A1a and A1b are arranged and stored in the order of candidates. Yes. The candidate characters stored in each record are stored together with a similarity that is an index representing the probability of the candidate characters recognized for the handwritten characters input in the respective character frames.

  For example, for the kana character “A” input by handwriting in the character frame A1a, the first candidate candidate character “O” is stored together with the similarity “3”. Similarly, the candidate character “a” of the second candidate is stored with the similarity “4”, and the candidate character “chi” of the third candidate is stored with the similarity “8”. Similarly, for the kana character “M” input by handwriting in the character frame A1b, the first candidate character “M” has the similarity “1” and the second candidate character “N” has the similarity “4”. In addition, the candidate character “a” of the third candidate is stored together with the similarity “9”.

  A correct candidate for a character handwritten in the character frame A1a is a second candidate with a certainty of similarity “4”, and a correct candidate for a character handwritten in the character frame A1b is a similarity “1”. It can be seen that the character is recognized as the first candidate with certainty.

  In the post-character recognition processing, for example, the candidate character group {O, A, Chi} recognized in the character frame A1a and the candidate character group {Me, N, A} recognized in the character frame A1b are brute-forced. A combination of two character strings is generated.

  FIG. 6C illustrates an explanatory diagram of a brute force combination of the candidate character group {O, A, C} and the candidate character group {Me, N, A}. As illustrated in FIG. 6C, the first candidate candidate character “O” in the character frame A1a includes the first to third candidate characters “Me”, “Nu”, and “A” in the character frame A1b. To be combined. In addition, the second candidate candidate character “A” in the character frame A1a is combined with the first to third candidate characters “ME”, “N”, and “A” in the character frame A1b. The third candidate character “chi” in the character frame A1a is combined with the first to third candidate characters “me”, “nu”, and “a” in the character frame A1b. As a result, nine character string groups are generated in the post-character recognition process.

  For example, the character recognizing unit 103 uses the word dictionary DB 201 for the character string group generated by combining the candidate character group recognized in the character frame A1a and the candidate character group recognized in the character frame A1b. Align to perform matching. For example, the character recognizing unit 103 adds the similarities for each candidate character included in each character string, calculates a similarity sum for each character string, and arranges the calculated similarities in order of magnitude.

  It is assumed that the similarity sum calculated from the similarity for each candidate character is the same value. When the similarity sums calculated for the candidate character strings are the same, for example, they can be arranged in the order of the character codes. For example, the character recognizing unit 103 temporarily stores a candidate character string group arranged in order of similarity sum and code order of character codes in a predetermined area of the main storage unit 12.

  FIG. 6D illustrates a configuration example of candidate character strings arranged in order of similarity sum. The configuration example in FIG. 6D has a record for each character string generated by combining the candidate character group recognized in the character frame A1a and the candidate character group recognized in the character frame A1b. Each record of the configuration example has columns of “candidate character string”, “sum of similarity”, “first character”, and “second character”. The “first character” and “second character” columns have a “character” column and a “similarity” column, respectively.

  The “candidate character string” column stores a character string obtained by combining candidate characters recognized in each character frame. In the “similarity sum” column, the similarity sum calculated for the character strings stored in the “candidate character string” column is stored. The “character” sub-column of the “first character” column stores candidate characters recognized in the character frame A1a. In the “character” sub-column of the “second character” column, candidate characters recognized in the character frame A1b are stored. The “similarity” sub-column of the “first character” column stores the similarity to the candidate character recognized in the character frame A1a. In the “similarity” sub column of the “second character” column, the similarity to the candidate character recognized in the character frame A1b is stored.

  In the configuration example of FIG. 6D, a value “5” calculated from the similarity of each character is stored in the “similarity sum” column of the record in which the character string “ame” as a correct answer candidate is stored. . Also, “a” of the first character of the character string is stored in the “character” sub-column of the “first character” column of the record, and the similarity “4” of “a” is stored in the “similarity” sub-column. Has been. In the same record, the “character” sub-column of the “second character” column stores “me” of the second character of the character string, and the “similarity” sub-column stores the similarity “1” of “me”. Has been.

  In the character recognition post-processing, for example, the character recognition unit 103 performs matching between each candidate character string arranged in order of similarity sum and code order of the character code and a registered word registered in the word dictionary DB 202. FIG. 6E illustrates an explanatory diagram of matching with a registered word registered in the word dictionary DB 202. In the explanatory diagram of FIG. 6E, the correct candidate character string “Ame” is modified by hatching.

  In the explanatory diagram of FIG. 6E, a registered word that is a two-character string is registered in the word dictionary DB 202. Registered words such as “Aki”, “Ashi”, “After”, “Ame”, “...” Registered in the word dictionary DB 202 are registered in the order of the code of the character code, for example.

  In the character recognition post-processing, for example, the character recognition unit 103 reads out the character string candidates stored in a predetermined area of the main storage unit 12 in the order of similarity sum order and character code order, and is registered in the word dictionary DB 202. Match with registered words.

  For example, the character recognizing unit 103 reads a candidate character string “Ome” having the lowest similarity sum (highest likelihood) and performs matching with a registered word registered in the word dictionary DB 202. The character recognition unit 103 reads a candidate character string having the next smallest similarity degree and registers it in the word dictionary DB 202 when there is no word that matches “Ome” among the registered words registered in the word dictionary DB 202. Repeat the matching with the registered word.

  In the example of FIG. 6E, the candidate character string “Ame” having the smallest similarity after the candidate character string “Ome” is read out. For example, when a word matching the read candidate character string “Ame” is registered in the word dictionary DB 202, the character recognition unit 103 temporarily stores the character string being processed in a predetermined area of the main storage unit 12. And finish the matching.

  In the character recognition post-processing, the character recognition unit 103, for example, in the candidate character group candidate order for the characters input in each character frame based on the candidate character string that matches the registered word registered in the word dictionary DB 202. Replace. For example, the character recognizing unit 103 exchanges the candidate characters stored in the character candidate buffer in the candidate order, which is handed over to the display control unit 104. The character candidate buffer in which the candidate character candidate ranks are switched is delivered to the display control unit 104.

  FIG. 6F illustrates a data example of the character candidate buffer in which the candidate rank is changed by the character recognition post-processing. In the example of FIG. 6F, the “candidate character” column of the “input box 1” column of the first candidate record is displayed in the “candidate character” column of the “input box 1” column of the second candidate record in the example of FIG. 6B. The stored candidate character “A” is stored. In the example of FIG. 6F, the “candidate character” column of the “input box 1” column of the second candidate record is displayed in the “candidate character” column of the “input box 1” column of the first candidate record in the example of FIG. 6B. The candidate character “o” stored in the column is stored. It can be seen that the candidate characters stored in the first candidate record and the candidate characters stored in the second candidate record illustrated in FIG. 6B are interchanged.

  In addition, in the first candidate record and the second candidate record illustrated in FIG. 6F, it can be seen that the similarity stored in the “similarity” sub-column is replaced along with the replacement of the candidate characters. For example, the character recognition unit 103 replaces the candidate ranks of the character candidates and replaces the similarities stored in the “similarity” subcolumn of the character candidate buffer.

FIG. 6G illustrates an example of a screen that is displayed after the candidate ranking is changed by the character recognition post-processing. As illustrated in FIG. 6G, “a”, which is the candidate character after replacement, is displayed in the display area A5 that displays the first candidate candidate character recognized for the handwritten character input in the character frame A1a. Yes. In addition, in the display area A6 that displays the second candidate candidate character recognized for the handwritten character input in the character frame A1a, the replaced candidate character “O” is displayed.

As a result of the post-character recognition process, “Ame”, which is a candidate character string that is a correct answer, is displayed in the first candidate for each character frame in which handwritten characters are input. In the example of FIG. 6G, for example, the display area A4 is displayed in a state in which “ Ame ” for the character string displayed as the first candidate is underlined to indicate an indeterminate state.

  Returning to the processing block diagram illustrated in FIG. 3, the display control unit 104 performs various controls of the display screen related to handwritten character input displayed on the LCD 15 a or the like, for example. For example, the display control unit 104 performs display control based on button operations on various operation buttons displayed on the display screen related to handwritten character input.

  In addition, the display control unit 104 displays handwriting of handwritten input characters on a display screen such as the LCD 15a based on the coordinate data stored in the coordinate data buffer by the coordinate input unit 102, for example. In addition, the display control unit 104 displays a candidate character group for handwritten characters input to each character frame as a recognition result based on, for example, data in the character candidate buffer delivered from the character recognition unit 103.

[Processing flow]
Hereinafter, with reference to the flowchart illustrated in FIGS. 7A to 7D, processing related to character recognition of handwritten input characters of the information processing apparatus 10 of the present embodiment will be described. The information processing apparatus 10 executes the processes illustrated in FIGS. 7A to 7D, for example, by a computer program that is developed in the main storage unit 12 so as to be executable.

  In the flowchart illustrated in FIG. 7A, the start of processing related to character recognition of handwritten input characters can be exemplified, for example, when an application that enables handwritten character input is activated. The information processing apparatus 10 detects, for example, a user operation with respect to an icon (icon) associated with an application that allows a handwritten character input displayed on a display screen such as the LCD 15a, and enables an input of a handwritten input character. Start up. For example, the information processing apparatus 10 performs an initialization process of a process related to character recognition of handwritten character input together with the activation of the application (S01).

  In the initialization process of S01, for example, the information processing apparatus 10 displays an initial screen related to handwritten character input illustrated in FIG. 2A or the like on the LCD 15a and the like, and performs initialization processes such as various buffers related to the process. Further, the information processing apparatus 10 starts time measurement by the timer 102a, for example. The initialization process has been described with reference to FIG.

  For example, the information processing apparatus 10 detects a contact position by a user operation on a display screen related to handwritten character input illustrated in FIG. 2A and the like displayed on a display device such as the LCD 15a (S02). For example, the information processing apparatus 10 detects a contact position by a user operation such as an operation finger on a display device surface such as the LCD 15a via an input device such as the touch sensor 14a. The contact position with respect to the display device surface is detected, for example, as coordinate data of (X, Y) two-dimensional coordinates with the upper left corner of the display device as the origin, the horizontal direction of the display device as the X axis, and the vertical direction as the Y axis. Is done.

For example, the information processing apparatus 10 determines whether the coordinate data of the contact position detected in the process of S02 is the display area in the input frame A1 illustrated in FIG. 2A or the like displayed on the display screen in the process of S01. Determine (S03). In the information processing apparatus 10, for example, the coordinate data of the contact position detected in the process of S02 is “input frame 1” − “input frame 3” of the coordinate table DB 203 illustrated in FIG. 5A.
It is determined whether or not it is within the coordinate area corresponding to any of “.” The character frames A1a to A1c in FIG. 2A and the like, and “input frame 1” to “input frame 3” in the coordinate table DB 203 illustrated in FIG. The relationship is described with reference to FIG. 5A.

  For example, the information processing apparatus 10 has a coordinate area in which the coordinate data of the contact position detected in the process of S02 corresponds to any one of “input frame 1” to “input frame 3” in the coordinate table DB 103 illustrated in FIG. 5A. If it is within (S03, Yes), the process proceeds to S10. On the other hand, the information processing apparatus 10 determines that the coordinate data of the contact position detected in the process of S02 is in any coordinate area of “input frame 1” to “input frame 3” in the coordinate table DB 203 illustrated in FIG. If not (S03, No), the process proceeds to S04 illustrated in FIG. 7B.

  In the flowchart illustrated in FIG. 7B, in the process of S04, the information processing apparatus 10 determines whether the button operation (pressing operation) is performed on the confirmation button A2 illustrated in FIG. For example, the information processing apparatus 10 determines whether or not the contact position detected in the process of S02 is within the coordinate area corresponding to the “confirm button” of the coordinate table 203 illustrated in FIG. 5A.

  For example, when the contact position detected in the process of S02 is within the coordinate area corresponding to the “confirm button” of the coordinate table 203 illustrated in FIG. 5A (S04, Yes), the information processing apparatus 10 The process proceeds to the process, and the character candidate string buffer illustrated in FIG. 5C is cleared.

  In the process of S06, the information processing apparatus 10 updates the display content of the display screen on which handwritten character input has been performed, for example, as illustrated in FIGS. 4A and 4B. Note that the updating of the display content by pressing the confirm button A2 has been described with reference to FIGS. 4A and 4B. For example, after updating the display content of the process of S06, the information processing apparatus 10 proceeds to the process of S02 illustrated in FIG. 7A and waits to detect the next contact operation.

  In the process of S04 in the flowchart illustrated in FIG. 7B, for example, the information processing apparatus 10 determines that the detected contact position is not within the coordinate area corresponding to the “confirm button” in the coordinate table 203 illustrated in FIG. 5A. (S04, No), the process proceeds to S07. In the process of S07, a pressing operation on the delete button A3 illustrated in FIG. 2A or the like is determined.

  In the process of S07, the information processing apparatus 10, for example, when the contact position detected in the process of S02 is within the coordinate area corresponding to the “delete button” of the coordinate table 203 illustrated in FIG. 5A (S07, Yes), the process proceeds to S08.

  In the process of S08, the information processing apparatus 10 stores, for example, the candidate character group stored in the character candidate string buffer for one character with respect to the input frame (character frame) input immediately before the delete button A3 is pressed. Erase the data. And in the process of S09, the information processing apparatus 10 updates the display content of the display screen in which the handwritten character input was performed, for example.

  On the updated display screen, for example, the handwriting of the handwritten character in the input frame that was input immediately before pressing the delete button A3 is deleted. Further, for example, the character candidate group displayed in the display area A5-A7 in association with the input frame input immediately before pressing the delete button A3 is deleted. In addition, for example, a character displayed in a state where character modification is applied to the display area A4, or a character for one character corresponding to an input frame input immediately before the delete button A3 is pressed from a character string. Erased. For example, after updating the display content of the process of S08, the information processing apparatus 10 proceeds to the process of S02 illustrated in FIG. 7A and waits to detect the next contact operation.

  Further, in the process of S07, the information processing apparatus 10, for example, when the contact position detected in the process of S02 is not within the coordinate area corresponding to the “delete button” of the coordinate table 203 illustrated in FIG. 5A ( (S07, No), the process again proceeds to the process of S02 illustrated in FIG. 7A.

  For example, when it is not an input operation or the like for each of the input frame A1, the confirm button A2, and the delete button A3 illustrated in FIG. 2A and the like, the contact position detected in S02 during processing is caused by an erroneous contact or the like. This is because it is considered to have occurred. For example, when the input operation is not performed on each of the input frame A1, the confirm button A2, and the delete button A3 illustrated in FIG. 2A and the like, the information processing apparatus 10 returns to the process of S02 and performs the next operation by the user operation or the like. Wait to detect the touch operation.

  Returning to the flowchart illustrated in FIG. 7A, in the process of S10, the information processing apparatus 10, for example, inputs the time-series coordinate data of the acquired contact position different from the input frame in which the contact position is determined in the process of S03. It is determined whether it is a frame area. For example, the information processing apparatus 10 acquires time-series coordinate data of the contact position. The information processing apparatus 10 determines the contact position in the process of S03 based on the relationship between the coordinate data of the contact position acquired in time series and the coordinate area corresponding to each input frame of the coordinate table DB 203, for example. It is determined whether or not the input frame is different from the input frame.

  For example, when the coordinate data of the contact position acquired in time series is included in the coordinate area of the input frame different from the input frame whose contact position is determined in the process of S03 (S10, Yes), the process proceeds to S14. On the other hand, for example, when the coordinate data of the contact position acquired in time series is included in the coordinate area of the input frame whose contact position has been determined in the process of S03 (S10, No), The process proceeds to S11.

  In the process of S11, the information processing apparatus 10 determines whether, for example, the measured value of the separation period of the operating finger or the like that has been in contact with the coordinate area of the input frame is equal to or greater than a predetermined threshold. For example, the information processing apparatus 10 detects the detachment of an operation finger or the like brought into contact with the display device surface from the acquired coordinate data of the time-series contact position. For example, the information processing apparatus 10 causes the timer 102a to function on the basis of the detected separation, measures the separation period until the coordinate data of the contact position is detected again, and determines the measured separation period and a predetermined threshold value. Make a comparison.

  For example, when the measured value of the measured withdrawal period is equal to or greater than a predetermined threshold (S11, Yes), the information processing apparatus 10 proceeds to the process of S14. On the other hand, for example, when the measured value of the measured withdrawal period is less than a predetermined threshold (S11, No), the information processing apparatus 10 proceeds to S12-S13.

  In the process of S12, the information processing apparatus 10 displays, for example, the time-series coordinate data of the acquired contact position as handwriting of characters input by handwriting in the corresponding input frame. For example, the information processing apparatus 10 stores the acquired coordinate data of the contact position in the coordinate data buffer illustrated in FIG. 5B (S13). For example, after the process of S13, the information processing apparatus 10 proceeds to the process of S02, and repeats the processes of S02-S03 and S10-S13.

  In the process of S14, for example, the information processing apparatus 10 delivers the time-series coordinate data of the contact position stored in the coordinate data buffer of the input frame in which the contact position is detected in the process of S03 to the character recognition process. The coordinate data stored in the coordinate data buffer includes a delimiter that distinguishes a stroke of a character input by handwriting in a character input frame to be processed. The character recognition process of the information processing apparatus 10 is executed by the character recognition unit 103, for example. The character recognition process executed by the character recognition unit 103 has been described with reference to FIG. 3, for example.

  In the process of S15, the information processing apparatus 10 clears, for example, time-series coordinate data serving as handwriting of handwritten input characters stored in the coordinate data buffer, and a measurement value (timer value) measured by the timer 102a. In the process of S16, the information processing apparatus 10 acquires, for example, the terminal state value accumulated in the acceleration sensor processing unit 103a. The acquired terminal state value is temporarily stored in a predetermined area of the main storage unit 12.

  FIG. 7D illustrates a flowchart of the terminal state value accumulation process executed by the acceleration sensor processing unit 103a. In the flowchart illustrated in FIG. 7D, the information processing apparatus 10 acquires, for example, acceleration values in the up / down / left / right and front / rear directions detected by the acceleration sensor 14a at regular intervals (S101). Here, as the fixed cycle interval, for example, a periodic interrupt cycle executed by the CPU 11 can be exemplified. The information processing apparatus 10 acquires the acceleration value detected by the acceleration sensor 14a by, for example, periodic interruption executed by the CPU 11.

  In the process of S102, for example, the information processing apparatus 10 compares the acceleration value acquired in the process of S101 with a predetermined threshold value, and determines whether the state (input environment) of the information processing apparatus 10 is a stable state or an unstable state. Determine. The result of the determination can be expressed as a flag value represented by a binary state of “0” and “1”, for example. For example, when the acquired acceleration value is equal to or greater than a predetermined threshold, the information processing apparatus 10 accumulates “1” representing an unstable state as a terminal state value. Further, for example, when the acquired acceleration value is less than a predetermined threshold, the information processing apparatus 10 accumulates “0” representing the stable state as the terminal state value.

  For example, the information processing apparatus 10 repeats the processing of S101 to S102 every time a periodic interrupt executed by the CPU 11 occurs. The terminal state value accumulation process has been described with reference to FIG.

  Returning to the flowchart illustrated in FIG. 7A, in the process of S <b> 17, the information processing apparatus 10 performs character recognition based on, for example, time-series coordinate data of the contact position handed over in the process of S <b> 14. For example, as a result of character recognition, the information processing apparatus 10 stores a candidate character group corresponding to the time-series coordinate data of the recognized contact position in the candidate character string buffer illustrated in FIG. In the character candidate string buffer, the similarity, which is an index value indicating the likelihood of the candidate character, is stored together with the candidate character.

  For example, the information processing apparatus 10 refers to the character recognition dictionary DB 201 and collates the time-series coordinate data of the contact position handed over in the processing of S14 with the registered characters registered in the character recognition dictionary DB 201. As a result of the collation, the information processing apparatus 10 specifies a candidate character group according to, for example, the degree of similarity with a registered character registered in the character recognition dictionary DB 201. For example, a candidate rank corresponding to the degree of similarity is given to the candidate character group. The specification of the character candidate group by the registered characters registered in the character recognition dictionary DB 201 has been described with reference to FIG.

  Returning to the flowchart illustrated in FIG. 7A, in the process of S18, the information processing apparatus 10 determines, for example, whether the time-series coordinate data of the contact position handed over in the process of S14 is the input of the first character. . For example, the information processing apparatus 10 determines that the time-series coordinate data of the contact position handed over in the process of S14 from the coordinate area associated with “input frame 1” in the coordinate table DB 203 illustrated in FIG. 5A is the first character. It is determined whether the input is.

  For example, when the time-series coordinate data of the contact position handed over in the process of S14 is input of the first character (S18, Yes), the information processing apparatus 10 performs the process of S26 in the flowchart illustrated in FIG. 7C. Migrate to

  For example, when the time-series coordinate data of the contact position handed over in the process of S14 is the input of the first character, the information processing apparatus 10 improves the character recognition rate by combining candidate character groups as a character string. Cannot perform post-character recognition processing. For this reason, the information processing apparatus 10 proceeds to the process of S26. For example, the first candidate character group stored in the character candidate string buffer in the process of S17 is used as the character recognition result for the input frame being processed. It is displayed on a display screen such as the LCD 15a.

  For example, as illustrated in FIG. 2A and the like, in the display area A5-A7 of the display screen such as the LCD 15a of the information processing apparatus 10, a candidate character group for the first character input frame includes candidate characters corresponding to the candidate ranks. Is displayed as a character recognition result. In addition, the candidate character recognized as the first candidate for the input frame of the first character is displayed in the display area A4 in a state in which character modification indicating an unconfirmed state is performed.

  Returning to the flowchart illustrated in FIG. 7A, in the process of S18, the information processing apparatus 10, for example, when the time-series coordinate data of the contact position handed over in the process of S14 is not the first character input (S18, No), the process proceeds to S19. In the process of S19, the information processing apparatus 10 determines whether the terminal state value acquired in the process of S16 is “1”, for example.

  For example, if the information processing apparatus 10 can determine that the input is not the first character in the process of S18, the character that improves the character recognition rate for the character string combined with the candidate character group recognized in another input frame. Post-recognition processing can be performed.

  Here, as described with reference to FIG. 2A, when the input environment of the information processing apparatus 10 related to handwritten character input is in a stable state, the correct candidate character tends to be recognized as the top candidate. . For example, with other candidate characters excluding the highest candidate, the degree of similarity indicating the probability of the candidate character tends to be low. For this reason, in a stable input environment, for example, a combination of candidate characters recognized as a top candidate is adopted without changing the candidate rank in the post-character recognition processing, so that the character recognition accuracy is high. It becomes possible to display a combination of candidate characters.

  On the other hand, as described with reference to FIG. 2B, when the input environment of the information processing apparatus 10 related to handwritten character input is in an unstable state, the ratio of candidate characters that are correct answers to the highest candidate is small, and the highest The similarity of the candidate character group including the higher candidates tends to be low. Therefore, in an unstable input environment, for example, post-character recognition processing is executed, and the candidate character group is switched based on the match with the registered word as a character string, so that the character recognition accuracy is high and correct. It becomes possible to display a combination of candidate characters.

  For example, the information processing apparatus 10 determines the input environment state of the information processing apparatus 10 related to handwritten character input based on the flag value of the terminal state value acquired in the process of S16, and thereby the character corresponding to the input environment state Post-recognition processing can be executed.

  In the process of S19 illustrated in FIG. 7A, the information processing apparatus 10 determines whether the terminal state value acquired in the process of S16 is “1”, for example. For example, when the terminal state value is “1”, the information processing apparatus 10 can determine that the input environment of the information processing apparatus 10 related to handwritten character input is in an unstable state. Similarly, for example, when the terminal state value is “0”, the information processing apparatus 10 can determine that the input environment of the information processing apparatus 10 related to handwritten character input is in a stable state.

For example, when the terminal state value acquired in the process of S16 is “0” (S19, No), the information processing apparatus 10 proceeds to the process of S26 in the flowchart illustrated in FIG. 7C. In the process of S26, for example, the information processing apparatus 10 displays the candidate character group stored in the character candidate string buffer in the process of S17 on the display screen such as the LCD 15a as a character recognition result for the input frame being processed.

  For example, as illustrated in FIG. 2A and the like, candidate characters corresponding to the candidate rank are displayed as character recognition results in the display area A5-A7 of the display screen such as the LCD 15a of the information processing apparatus 10. In the display area A4, a character string that is a combination of candidate characters recognized as the first candidate in each input frame is displayed with a predetermined character modification. In the display area A4, as described with reference to FIG. 2A and the like, a combination of candidate characters with a high probability of being correct is displayed as a character string before confirmation.

  Returning to the flowchart illustrated in FIG. 7A, in the process of S19, the information processing apparatus 10, for example, when the terminal state value acquired in the process of S16 is “1” (S19, Yes), the process illustrated in FIG. The process proceeds to S20 to S26 in the flowchart. For example, the information processing apparatus 10 performs the processing of S20 to S26 and performs matching between a character string generated from a combination of candidate character groups stored in the character candidate string buffer and a registered word registered in the word dictionary DB 201. And post-character recognition post-processing for exchanging candidate ranks. The information processing apparatus 10 combines a plurality of candidate characters that have been character-recognized for each handwritten input character with respect to a series of handwritten character inputs input to the touch panel, and executes the character recognition post-processing shown in S20 to S26.

  In the flowchart illustrated in FIG. 7C, in the process of S20, the information processing apparatus 10 combines, for example, each of the candidate character groups for the individual input frames stored in the character candidate string buffer, and sets all candidate character strings. Is generated. Each candidate character string generated by the process of S20 is associated with, for example, a similarity sum obtained by adding the similarities of individual candidate characters. For example, the information processing apparatus 10 arranges the generated candidate character strings in order of the similarity sum, and temporarily stores them in a predetermined area of the main storage unit 12. The generation of all candidate character strings in S20 has been described with reference to FIGS. 6B, 6C, and 6D, for example.

  In the process of S21, the information processing apparatus 10, for example, selects a candidate character string that is a matching target from all candidate character strings generated in the process of S20. For example, the information processing apparatus 10 extracts candidate character strings to be matched in the order of arrangement from the generated candidate character strings.

  In the process in S22, the information processing apparatus 10 determines whether matching with all candidate character strings generated in the process in S20 is completed, for example. For example, the information processing apparatus 10 determines that there is no candidate character string to be matched in all candidate character strings stored in a predetermined area of the main storage unit 12.

  For example, when there is no candidate character string to be matched in all candidate character strings stored in a predetermined area of the main storage unit 12 (S22, Yes), the information processing apparatus 10 proceeds to A26. . In the process of S26, for example, the information processing apparatus 10 displays the candidate character group stored in the character candidate string buffer in the process of S17 on the display screen such as the LCD 15a as a character recognition result for the input frame being processed. On the display screen such as the LCD 15a, the candidate character groups recognized in the individual input frames are displayed in a predetermined display area without changing the candidate rank.

On the other hand, for example, when there is no candidate character string to be matched in all candidate character strings stored in a predetermined area of the main storage unit 12 (S22, No), the information processing apparatus 10 proceeds to S23. Transition. In the process of S23, the information processing apparatus 10 uses, for example, the word dictionary DB 20
2, matching is performed between the candidate character string extracted in the process of S <b> 21 and the registered word registered in the word dictionary DB 202. The result of matching between the candidate character string extracted in the process of S21 and the registered word registered in the word dictionary DB 202 is determined in the process of S24.

  In the process of S24, the information processing apparatus 10 determines whether or not a registered word that matches the candidate character string extracted in the process of S21 exists in the word dictionary DB 202. For example, if the registered word that matches the candidate character string extracted in the process of S21 does not exist in the word dictionary DB 202 as a result of the process of S23 (S24, No), the information processing apparatus 10 proceeds to the process of S21. Then, the processing of S21-S24 is repeated for the next candidate character string.

  On the other hand, for example, when a registered word that matches the candidate character string extracted in the process of S21 exists in the word dictionary DB 202 (S24, Yes), the information processing apparatus 10 proceeds to the process of S25.

  In the process of S25, the information processing apparatus 10, for example, based on the registered word that matches the candidate character string extracted in the process of S21, the individual input stored in the character candidate string buffer in the process of S17 of FIG. 7A. The candidate ranks of the candidate character groups recognized in the frame are switched. Note that the replacement of candidate ranks for candidate character groups recognized in each input frame and stored in the character candidate string buffer has been described with reference to FIGS. 6A, 6B, 6F, and 6G, for example.

  For example, for the input of the handwritten character “Ame”, “O” as the first candidate, “A” as the second candidate, and “Chi” as the third candidate are recognized as the candidate characters for the first character “A”. Assuming that Further, it is assumed that “me” as the first candidate, “nu” as the second candidate, “a” as the third candidate, and “a” as the candidate character are recognized as the candidate characters. For example, as illustrated in FIG. 6A, in the character string displayed in the display area A4, “Ome”, which is a combination of the candidate characters of the first candidate, is displayed as the character string of the character recognition result.

  For example, when “ame” is registered as a registered word in the word dictionary DB 202, the information processing apparatus 10 determines that the first candidate “O” and the second candidate “A” are the first character. Swap candidate ranks. As a result, for example, as illustrated in FIG. 6F, the second candidate “a” having a relatively low degree of similarity becomes the candidate character of the first candidate, and the first candidate “O” having a relatively high degree of similarity is used. Is stored in the character candidate string buffer as the candidate character of the second term.

  Returning to the flowchart illustrated in FIG. 7C, in the process of S26, the information processing apparatus 10, for example, based on the candidate character group stored in the candidate character string buffer in which the candidate ranking is changed by the post-character recognition process, Displays the character recognition result.

For example, as illustrated in FIG. 6G, the display area A5 in which the candidate character is displayed as the first candidate for the first character is displayed as “a” after replacement, and the candidate character is displayed as the second candidate. In the area A6, “O” after the replacement is displayed. In addition, in the display area A4, a character string “ Ame ” that is a combination of candidate characters recognized as the first candidate for the first character and the second character has been subjected to character modification with an underline indicating the state before confirmation. Displayed with status. As a result of the character recognition post-processing of S20 to S26, the information processing apparatus 10 registers a match with a combination of a plurality of candidate characters recognized for each handwritten input character with respect to a series of handwritten character inputs input to the touch panel. Candidate character display based on words can be performed.

  Returning to the flowchart illustrated in FIG. 7C, after the process of S <b> 26, the information processing apparatus 10 returns to the process of S <b> 02 of the flowchart illustrated in FIG. 7A, for example, and waits to detect the next contact operation due to a user operation or the like. To do.

  Here, the process of S02-S14 performed by the information processing apparatus 10 is an example of a character recognition step of recognizing a plurality of candidate characters from the contact coordinates detected by the touch panel. Further, the CPU 11 or the like of the information processing apparatus 10 executes the processes of S02 to S14 as an example of a character recognition unit that recognizes a plurality of candidate characters from the contact coordinates detected by the touch panel.

  Moreover, the process of S19 performed with the information processing apparatus 10 is an example of the step which determines the motion of the housing | casing containing a touch panel. Further, the CPU 11 or the like of the information processing apparatus 10 executes the process of S19 as an example of a unit that determines the movement of the casing including the touch panel.

  Further, in the processing of S20 to S26 executed by the information processing apparatus 10, when it is determined that the movement of the housing is equal to or greater than a predetermined amount, a word dictionary is created by combining a plurality of candidate characters recognized in the character recognition step. It is an example of the step which searches and outputs the candidate character string based on a search result. Further, the CPU 11 or the like of the information processing apparatus 10 searches the word dictionary with a combination of a plurality of candidate characters recognized by the character recognition unit when the movement of the housing is determined to be a predetermined amount or more, and the search result is As an example of a means for outputting a candidate character string based on it, the processing of S20-S26 is executed.

  Further, in the processing of S19, “No” -S26 executed by the information processing apparatus 10, the character recognition step calculates the similarity indicating the degree of matching between the candidate character and the character shape that is the character recognition reference, When it is determined that the movement of the housing is less than a predetermined amount, this is an example of a step of outputting a plurality of candidate characters recognized in the character recognition step in the order of similarity calculated for each candidate character. In addition, the CPU 11 or the like of the information processing apparatus 10 calculates the similarity indicating the degree of coincidence between the candidate character and the character shape serving as a character recognition reference, and determines that the movement of the housing is less than the predetermined amount. If it is, the process of S19, “No” -S26 is executed as an example of a means for outputting a plurality of candidate characters recognized by the character recognition means in the order of similarity calculated for each candidate character.

  As described above, in the information processing apparatus 10 according to the present embodiment, the state of the input environment of the information processing apparatus 10 related to handwritten character input can be determined based on the detection value detected by the acceleration sensor 14a or the like. it can. In the information processing apparatus 10, for example, when the detection value detected by the acceleration sensor 14 a or the like is equal to or greater than a predetermined threshold, the input environment of the information processing apparatus 10 related to handwritten character input can be determined to be in an unstable state.

  When the information processing apparatus 10 determines that the input environment of the information processing apparatus 10 related to handwritten character input is in an unstable state, the information processing apparatus 10 refers to the word dictionary DB 202 and stores candidate character groups recognized in individual input frames. It is possible to perform post-character recognition processing for changing the candidate rank. As a result of the character recognition post-processing, the information processing apparatus 10 can increase the character recognition rate of handwritten characters input to each input frame based on the character strings of registered words registered in the word dictionary DB 202.

  In addition, when it is determined that the input environment of the information processing apparatus 10 related to handwritten character input is in a stable state, the information processing apparatus 10 according to the present embodiment includes, for example, candidate character groups recognized in individual input frames. It is possible to display based on an index representing the certainty.

  For this reason, the information processing apparatus 10 of this embodiment can reduce, for example, the processing load on the CPU 11 and the operation rate related to the post-character recognition processing. Moreover, the information processing apparatus 10 of this embodiment can reduce the processing time of the character recognition result about the handwritten character input into each input frame in the case of a stable input environment. As a result, in the information processing apparatus 10 of the present embodiment, it is possible to suppress power consumption related to character recognition of the information processing apparatus 10. In the information processing apparatus 10 according to the present embodiment, it is possible to realize a technique for suppressing power consumption according to an environmental state at the time of handwriting input of an electronic device.

<Modification>
(1) In the embodiment, for example, all candidate characters recognized as candidate characters (three in the embodiment) are set as targets for post-character recognition processing. In the character recognition post-processing as an example of the modification, for example, the number of candidate characters to be subjected to the character recognition post-processing may be determined based on the probability (similarity) obtained for each candidate character.

  For example, the information processing apparatus 10 provides a selection threshold for selecting a candidate character as a processing target of post-character recognition processing for the similarity obtained for each candidate character. Then, for example, in the process of S20 of the flowchart illustrated in FIG. 7C, the information processing apparatus 10 compares the similarity of each candidate character stored in the character candidate string buffer with a selection threshold, and after character recognition Select a candidate character to be processed.

  For example, taking the character candidate string buffer in FIG. 5C as an illustrative example, when the selection threshold for similarity is “8”, a candidate character storing a similarity of less than 8 is selected as a target for post-character recognition processing. be able to. For example, from the character candidate string buffer shown in FIG. 5C, select two characters {a, o} in input frame 1, {Shi, ku} in input frame 2, and {ta, ni} in input frame 3. Can do. In this case, the number of candidate character strings to be generated is “2 types × 3 types × 2 types = 12 types”, and the processing time and processing load on the CPU 11 related to the post-character recognition processing performed on the 27 candidate character strings. The operation rate can be relatively reduced. As a result, further reduction of power consumption can be expected.

  For example, the selection threshold value may be varied according to the detection value detected by the acceleration sensor 14a. For example, when the movement and shaking of the information processing apparatus 10 are relatively large, the selection threshold for the similarity may be changed to a low value such as “8” → “11”. Similarly, for example, when the movement and shaking of the information processing apparatus 10 are relatively small, the selection threshold for the similarity may be changed to a high value such as “8” → “5”.

  Depending on the degree of movement of the information processing apparatus 10, the processing time for post-character recognition processing, the processing load on the CPU 11, the operating rate, and the like can be relatively reduced. It can be expected that power consumption is suppressed according to the degree of movement of the information processing apparatus 10.

  Furthermore, the application target of the selection threshold value can be the similarity sum of candidate character strings generated by combining candidate characters. For example, referring to FIG. 6D, when the selection threshold for the similarity sum is “11”, a candidate character string in which the similarity sum less than 11 is stored as a registered word registered in the word dictionary DB 202. It can be selected as a character string related to the matching process. By limiting the number of candidate character strings related to the matching process with the sum of similarities, it is possible to relatively reduce the processing time related to the matching process, the processing load on the CPU 11, the operating rate, and the like.

(2) In the embodiment, the execution timing of the character candidate post-processing is executed for each handwritten character input after the second character. In the post-character recognition processing as an example of the modification, for example, character recognition may be performed on all input frames displayed on the display screen, and then may be executed collectively. In the example of FIG. 2A, for example, three characters obtained by combining candidate character groups for each character frame after all candidate character groups for handwritten characters input to the character frames A1a-A1c are recognized. The character recognition post-processing may be performed collectively for the character strings. After the character recognition is performed on all the input frames displayed on the display screen, the character recognition reflecting the intention of the user who has input the handwritten character can be expected by collectively executing the post-character recognition process.

(3) In the embodiment, in the stable state, A20-S25, which is the post-character recognition process, is skipped, and the candidate character groups recognized for the character are displayed in the order of candidates based on the similarity. In the post-character recognition process as an example of the modification, for example, the post-character recognition process may be executed even in a stable state. However, the post-character recognition post-processing of the modified example is executed by limiting the number of candidate characters to be subjected to post-character recognition processing. For example, as described in the modification example (1), it is possible to exemplify limiting the number of processing targets of the character recognition post-processing by the similarity of candidate characters.

  For example, in the modification (3), a first selection threshold having a higher similarity than the selection threshold described in the modification (1) can be provided as a selection threshold for candidate characters to be subjected to post-character recognition processing. In the information processing apparatus according to the modified example (3), for example, even when it is determined to be in a stable state, a candidate character having a higher similarity than the first selection threshold value is selected, and character recognition post-processing is performed. . In the information processing apparatus of the modification (3), the character recognition post-processing described in the modification (1) can be performed. In this case, for example, the selection threshold for selecting candidate characters for post-character recognition processing executed in a stable state is set as the first selection threshold, and the second selection threshold having a similarity lower than the first selection threshold is not set. It can be set as a selection threshold for selecting candidate characters for post-character recognition processing executed in a stable state. The magnitude relationship between the first selection threshold and the second selection threshold can be obtained experimentally.

  Further, for example, in the example of FIG. 2A, the candidate character group that is a character recognition target can be limited to character attributes such as kana characters. In addition, for example, a user dictionary unique to the user who uses the information processing apparatus 10, a dictionary in which “words and words” according to time and the like are registered, and the words registered in the above-described dictionary in a stable state are provided. May be processed preferentially.

<Computer-readable recording medium>
A program for causing a computer or other machine or device (hereinafter, a computer or the like) to realize any of the above functions can be recorded on a recording medium that can be read by the computer or the like. The function can be provided by causing a computer or the like to read and execute the program of the recording medium.

  Here, a computer-readable recording medium is a recording medium that stores information such as data and programs by electrical, magnetic, optical, mechanical, or chemical action and can be read from a computer or the like. Say. Examples of such a recording medium that can be removed from a computer or the like include a flexible disk, a magneto-optical disk, a CD-ROM, a CD-R / W, a DVD, a Blu-ray disk, a DAT, an 8 mm tape, a flash memory, and the like. There are cards. Moreover, there are a hard disk, a ROM, and the like as a recording medium fixed to a computer or the like.

<Others>
The above embodiment further includes an aspect called the following supplementary note. The components included in the following supplementary notes can be combined with the constituents included in the other supplementary notes.
(Appendix 1)
Character recognition means for recognizing a plurality of candidate characters from contact coordinates detected by the touch panel;
Means for determining the movement of the housing including the touch panel;
Means for searching a word dictionary with a combination of a plurality of candidate characters recognized by the character recognition means and outputting a candidate character string based on the search result when it is determined that the movement of the housing is equal to or greater than a predetermined amount; When,
Electronic equipment comprising. (1)
(Appendix 2)
The character recognition means calculates a similarity indicating a degree of matching between the candidate character and a character shape as a reference for character recognition;
The apparatus further includes means for outputting a plurality of candidate characters recognized by the character recognition means in order of similarity calculated for each of the candidate characters when it is determined that the movement of the housing is less than a predetermined amount. 1. The electronic device according to 1. (2)
(Appendix 3)
The character recognition means calculates a similarity indicating a degree of matching between the candidate character and a character shape as a reference for character recognition;
If it is determined that the movement of the housing is less than a predetermined amount, the similarity calculated for the candidate character is greater than or equal to a first threshold value from among a plurality of candidate characters recognized by the character recognition means. The electronic device according to appendix 1 or 2, further comprising means for selecting a candidate character as a processing target, searching a word dictionary with the selected combination of candidate characters, and outputting a candidate character string based on the search result . (3)
(Appendix 4)
If it is determined that the movement of the casing is equal to or greater than a predetermined amount, the similarity calculated for the candidate character from the plurality of candidate characters recognized by the character recognition means is greater than the first threshold value. The electronic device according to appendix 3, wherein a candidate character having a lower second threshold value or higher is selected as a processing target, a word dictionary is searched with a combination of the selected candidate characters, and a candidate character string based on the search result is output. machine. (4)
(Appendix 5)
The said character recognition means varies the 1st threshold value for selecting the candidate character used as a candidate for processing from a plurality of recognized candidate characters according to the degree of the movement of the above-mentioned case, The additional statement 3 Electronics. (5)
(Appendix 6)
The said character recognition means varies the 2nd threshold value for selecting the candidate character used as a process target from several recognized candidate characters according to the grade of the movement of the said housing | casing, Additional statement 4 Electronics. (6)
(Appendix 7)
On a computer with a word dictionary,
A character recognition step for recognizing a plurality of candidate characters from the contact coordinates detected by the touch panel;
Determining the movement of the housing including the touch panel;
A step of searching a word dictionary with a combination of a plurality of candidate characters recognized in the character recognition step and outputting a candidate character string based on the search result when it is determined that the movement of the housing is equal to or greater than a predetermined amount When,
Handwritten character input program to execute.
(Appendix 8)
The character recognition step calculates a similarity indicating a degree of coincidence between the candidate character and a character shape serving as a character recognition reference,
If it is determined that the movement of the housing is less than a predetermined amount, a step of outputting the plurality of candidate characters recognized in the character recognition step in the order of similarity calculated for each of the candidate characters is further executed. The handwritten character input program according to appendix 7.
(Appendix 9)
The character recognition step calculates a similarity indicating a degree of coincidence between the candidate character and a character shape serving as a character recognition reference,
If it is determined that the movement of the housing is less than a predetermined amount, a similarity calculated for the candidate character is greater than or equal to a first threshold value from among a plurality of candidate characters recognized in the character recognition step. The handwriting according to appendix 7 or 8, wherein a candidate character is selected as a processing target, and a step of searching a word dictionary with the selected candidate character combination and outputting a candidate character string based on the search result is further executed. Character input program.
(Appendix 10)
If it is determined that the movement of the housing is equal to or greater than a predetermined amount, the similarity calculated for the candidate character from the plurality of candidate characters recognized in the character recognition step is the first threshold value. The candidate character string according to claim 9, wherein candidate characters that are lower than the second threshold value are selected as processing targets, a word dictionary is searched with the combination of the selected candidate characters, and a candidate character string based on the search result is output. Handwritten character input program.
(Appendix 11)
The said character recognition step changes the 1st threshold value for selecting the candidate character used as a process target from several recognized candidate characters according to the grade of the motion of the said housing | casing, The additional statement 9 Handwritten character input program.
(Appendix 12)
The character recognition step according to appendix 10, wherein the second threshold value for selecting a candidate character to be processed from among a plurality of recognized candidate characters is varied according to the degree of movement of the housing. Handwritten character input program.

10 Information processing equipment (electronic equipment)
11 CPU
12 Main storage unit 13 Auxiliary storage unit 14 Input unit 14a Touch sensor 14b Acceleration sensor 15 Output unit 15a LCD
16 Communication unit 101 Initial setting unit 102 Coordinate input unit 102a Timer 103 Character recognition unit 103a Acceleration sensor processing unit 104 Display control unit 201 Character recognition dictionary DB
202 Word Dictionary DB
203 Coordinate table DB
204 Screen image DB

Claims (7)

Character recognition means for recognizing a plurality of candidate characters from contact coordinates detected by the touch panel;
Means for determining the movement of the housing including the touch panel;
Means for searching a word dictionary with a combination of a plurality of candidate characters recognized by the character recognition means and outputting a candidate character string based on the search result when it is determined that the movement of the housing is equal to or greater than a predetermined amount; When,
Electronic equipment comprising. The character recognition means calculates a similarity indicating a degree of matching between the candidate character and a character shape as a reference for character recognition;
And a means for outputting a plurality of candidate characters recognized by the character recognition means in order of similarity calculated for each of the candidate characters when it is determined that the movement of the housing is less than a predetermined amount. Item 2. The electronic device according to Item 1. The character recognition means calculates a similarity indicating a degree of matching between the candidate character and a character shape as a reference for character recognition;
If it is determined that the movement of the housing is less than a predetermined amount, the similarity calculated for the candidate character is greater than or equal to a first threshold value from among a plurality of candidate characters recognized by the character recognition means. The electronic device according to claim 1, wherein a candidate character is selected as a processing target, a word dictionary is searched with the selected combination of candidate characters, and a candidate character string based on a search result is output.   If it is determined that the movement of the casing is equal to or greater than a predetermined amount, the similarity calculated for the candidate character from the plurality of candidate characters recognized by the character recognition means is greater than the first threshold value. The candidate character more than the 2nd threshold value which is lower than 2 is selected as a candidate for processing, a word dictionary is searched with the combination of the selected candidate characters, and a candidate character string based on a search result is outputted. Electronics.   The said character recognition means changes the 1st threshold value for selecting the candidate character used as a process target from several recognized candidate characters according to the grade of the motion of the said housing | casing. Electronic equipment.   The said character recognition means changes the 2nd threshold value for selecting the candidate character used as a process target from several recognized candidate characters according to the grade of the motion of the said housing | casing. Electronic equipment. On a computer with a word dictionary,
A character recognition step for recognizing a plurality of candidate characters from the contact coordinates detected by the touch panel;
Determining the movement of the housing including the touch panel;
A step of searching a word dictionary with a combination of a plurality of candidate characters recognized in the character recognition step and outputting a candidate character string based on the search result when it is determined that the movement of the housing is equal to or greater than a predetermined amount When,
Handwritten character input program to execute.

Images