JP4381632B2 - Navigation system and its destination input method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a navigation device and a destination input method thereof, and more particularly to a navigation device and a destination input method for inputting a destination by key input and voice input.
[0002]
[Prior art]
When the destination is input, the navigation device searches for a route to the destination and guides the driver along the route. To enter a destination,
(1) Key-in destination address (address input),
(2) If the destination is an intersection, specify two streets and input the intersection (intersection input),
(3) Key input method of POI (Point of Interest) name (POI input)
(4) A method for displaying a POI candidate list by inputting a category and a city name, and selecting a destination POI from the candidate list,
(5) Directly enter the destination on the map with the cursor,
(6) A method in which a predetermined name is assigned to an address and registered in the address book, and the address book is displayed and the destination address is input.
(7) Enter the POI phone number,
and so on.
As a recent trend, a destination is specified by inputting a voice instead of a key.
[0003]
[Problems to be solved by the invention]
Voice input is convenient because it does not require key operation compared to key input. However, in order to perform speech recognition, it is necessary to register a word to be recognized in the speech recognition dictionary in advance. In the method of inputting the destination by address, it is necessary to input the city, street, etc. by voice, and when inputting by POI, it is necessary to input the POI name (facility name) by voice. There are so many names, hundreds of thousands. For this reason, in order to correctly recognize the destination, there is a problem that the speech recognition dictionary is too large, the hardware scale is increased, the apparatus is increased in size, and the cost is increased.
Accordingly, an object of the present invention is to enable correct speech recognition of a destination with a small hardware scale.
[0004]
[Means for Solving the Problems]
A first aspect of the present invention is a destination input method of a navigation apparatus for inputting a destination by key input and voice input. In the destination input mode, a name of a region to which the destination belongs or a name of a facility as a destination is keyed. And processing for specifying the destination using the area name or facility name input by the key, and synthesizing the voice of the area name or facility name input by the key, In the destination input mode, the name of the area to which the destination belongs or the name of the facility as the destination is input by voice with reference to the voice dictionary. Identify the name of the area to which the destination belongs, or the name of the facility as the destination, and specify the destination using the name of the area to which the identified destination belongs or the name of the facility as the destination. A process to perform.
By doing so, it is only necessary to add and register only the voice pattern of the area name and facility name corresponding to the destination required by the user to the initial registered voice pattern, so the number of voice patterns to be registered This makes it possible to recognize the destination correctly with a small hardware scale.
[0005]
In addition, when inputting a destination by gradually narrowing from an upper area of a large area to a lower area of a small area, the area name of the lower area belonging to the upper area and its synthesized speech are associated with the upper area. may be registered, when the upper region in which it has been the voice input has been identified, to recognize the sub-region in which it has been voice input with reference to the regional name and its synthetic speech of the lower region belonging to the upper level area in the next voice input To. In this way, the range of searching for a speech pattern similar to the input speech is narrowed, and speech recognition can be performed with high accuracy in a short time.
[0006]
A second aspect of the present invention is a navigation system for inputting a destination by key input and voice input, key input means for key input of a name of a region to which the destination belongs or a facility name as the destination in the destination input mode, A speech synthesizer that synthesizes the voice of the area name or facility name that is keyed in, a speech dictionary part that registers the synthesized speech corresponding to the area name or facility name, a facility as the area name or destination to which the destination belongs When a name is input by voice, by referring to the voice dictionary unit, a voice recognition unit for identifying a name of a region to which the destination input by voice belongs or a facility name as a destination, in a destination input mode, by a key or Navigation that performs processing to identify the destination using the name of the area to which the destination belongs or the name of the facility as the destination It is equipped with a control unit. According to this navigation system, the number of voice patterns to be registered can be reduced, and the destination can be correctly voice-recognized with a small hardware scale.
[0007]
DETAILED DESCRIPTION OF THE INVENTION
(A) Navigation Device (a) Configuration FIG. 1 is a configuration diagram of a navigation device 10 of the present invention. In the figure, 11 is a map storage medium for storing map information, for example, a DVD (digital video disc), 12 is a DVD control unit for controlling reading of map information from the DVD, and 13 is a position measuring device for measuring the current vehicle position. The vehicle speed sensor for detecting the moving distance, the angular velocity sensor for detecting the moving direction, the GPS receiver, the position calculating CPU, and the like. 14 is a map information memory for storing map information around the vehicle position read from the DVD, 15 is a remote controller for performing various settings and instructions by selecting a menu, inputting a point name, and enlarging / reducing, etc. 16 is a remote control interface It is.
[0008]
Reference numeral 17 denotes a CPU (navigation control unit) for controlling the entire navigation apparatus, 18 denotes software (loading program) for downloading various control programs from the DVD, ROM for storing fixed data, etc., 19 denotes various types downloaded from the DVD The RAM stores a control program (destination setting program DSP, route search program RSP, etc.), searched guide route data RT, and other processing results. Reference numeral 21 denotes a storage unit for storing an address database read from a DVD. A city is listed for each state, a street is listed for each city, and a history of a point indicated by an address (state / city / street / address). Also has a degree.
[0009]
A storage unit 22 stores a POI database read from a DVD, and holds (1) address, (2) category, (3) POI position (longitude and latitude) for each POI name. 23 is a display controller for generating a map image / guidance route, 24 is a video RAM for storing an image generated by the display controller, 25 is a menu / list generating unit for generating various menus and lists, and 26 is for combining various images. 27 is a display device (monitor) that displays an image output from the image composition unit, 28 is a voice guidance unit that guides the distance to the intersection and the traveling direction by voice, and 29 is a communication interface. A device 30 transmits and receives data to and from the voice recognition device, and 30 is a bus.
[0010]
(B) Destination Identification Method by Destination Input Method / Address Input FIGS. 2 and 3 are explanatory diagrams of a destination specification method by address input. US address notation has the format state / city name / street name / address. Therefore, the destination can be specified by inputting the city name / street name / address. Therefore, in order to specify the destination by inputting an address in the navigation device, the user operates the menu key of the remote controller to display the main menu on the screen, and selects the menu item “Dest” from the main menu. By selecting “Dest”, the navigation device displays the “Find Destination by” screen for specifying the destination input method as shown in FIG. 2 (a). When the enter key is pressed, the navigation control unit displays the menu items “Change State”, “City Name”, and “Street Name” as shown in (b). To enter the destination address starting from the city name, select the menu item “City Name”, and to enter the destination address starting from the street name, enter the menu item “Street Name”. Also, since the last selected state is displayed as the default state, if you want to change the state name, select "Change State" ((c)) and scroll to select the desired state ((d )).
[0011]
After the state is determined, when the menu item “City Name” is selected and the enter key is operated, the navigation control unit displays an alphabet / numeric keyboard for inputting the city name on the screen as shown in (e). FIG. 4 shows an enlarged keyboard image. In this state, when the user inputs the first spelling of the city name and inputs the list key List on the keyboard ((f)), as shown in FIG. 3 (g), the city having the spell in California A list of names is displayed in a scrollable manner. The figure shows the case where the spelling “T” is entered, and the city name “TORRANCE” having “T” at the head is displayed. Next, use the joystick key to scroll the city name list, select the desired city name (TORRANCE), and operate the enter key.
[0012]
As described above, when the city name specifying operation is completed, the navigation device displays a street name input screen as shown in (h). In this state, the user inputs the first few spells of the street and presses the list key List on the keyboard ((i)). Thereby, as shown in (j), a list of street names existing in the TORRANCE having the spelling is displayed in a scrollable manner. The figure shows a case where the spelling “CAR” is entered, and a street name “CARSON” having “CAR” at the head is listed. Next, use the joystick key to scroll the street name list, select a desired street name (CARSON), and operate the enter key. As described above, when the street name specifying operation is completed, the navigation control unit displays the address input screen as shown in (k). In this state, if the address is input and the enter key is operated, the destination address is displayed as shown in (m). If the address is correct as the destination, select "OK to Proceed" to confirm the destination.
When the destination address is input, the navigation control unit obtains the longitude and latitude of the destination address from the address database 21 and performs a route search using the position information.
[0013]
-Specifying the destination by entering an intersection To specify two streets and input those intersections as destinations, select the menu item "Intersction" on the "Find Destination by" screen in Fig. 2 (a). Thereafter, the destination can be specified by inputting the name of the city where the destination (intersection) is located and two streets in the same manner as the address input.
[0014]
-Destination identification by POI name input FIG. 5 is an explanatory diagram of a destination identification method by POI name input. In the navigation device, in order to specify the destination by inputting the POI name, the user operates the menu key of the remote controller to display the main menu on the screen, and selects the menu item “Dest” from the main menu. By selecting "Dest", the navigation controller displays the "Find Destination by" screen for specifying the destination input method as shown in Fig. 5 (a). Select the menu item "Point of Interest". If you press the Enter key on the remote control, the navigation device displays the menu items "Place Name", "Place Type sort by distance", and "Place Type within a city" as shown in (b). Select “Place Name” and press the Enter key. As a result, the navigation control unit displays an alphabet / numeric keyboard for inputting the POI name on the screen as shown in (c). In this state, the user inputs the first few spellings of the POI name, and then inputs the list key List on the keyboard as shown in (d).
[0015]
Thereby, as shown in (e), the navigation device displays a scrollable list of POI names throughout the United States having the spell. The figure shows the case where the spelling “DEL AMO” has been entered, and the navigation device displays the POI name having “DEL AMO” at the beginning. Also, the navigation device obtains the address, category (SHOPPING), and telephone number of the temporarily selected “DEL AMO FASHION CENTER” by referring to the POI database 22 and displays the destination direction on the screen. Next, use the joystick key on the remote control to scroll through the POI name list, select the desired POI (DEL AMO FASHION CENTER) and operate the enter key, the navigation device will display the destination address as shown in (f). Because it displays. If the address is correct as the destination, select "OK to Proceed" to confirm the destination. When the destination address is input, the navigation device obtains the longitude and latitude of the POI from the POI database 22 and performs a route search using the position information.
[0016]
(B) Voice Recognition Device FIG. 6 is a block diagram of the voice recognition device 50 of the present invention. In the figure, 51 is a microphone for detecting speaker speech, 52 is a speech dictionary database for storing word character strings and word speech patterns in association with each other, and 53 is a phoneme pattern database for storing phonemes and phoneme patterns in association with each other. is there. A phoneme is a part consisting of three alphabets when a character string is written in Roman letters. For example, as shown in FIG. SA, (2) SAP, (3) APP, (4) PPO, (5) POR, (6) ORO, (7) RO Becomes phoneme. An unspecified number of speakers utter many words, each speech pattern is decomposed for each phoneme, the phoneme pattern is obtained by averaging the same phoneme pattern, and the phoneme pattern is associated with the phoneme. Register with. Reference numeral 54 denotes a destination input voice dictionary for storing synthesized city names, street name voice patterns, and the like. Reference numeral 55 denotes a communication interface unit for exchanging data with the navigation device.
[0017]
Reference numeral 56 denotes a voice recognition engine, which includes a voice recognition unit 56a and a voice dictionary creation unit 56b. The voice recognition unit 56a searches for and outputs a character string having a voice pattern most similar to the voice input from the microphone by phoneme pattern matching. When the destination is input by key input in the navigation device, the voice dictionary creating unit 56b synthesizes the voice pattern of the input region (state, city) name, street name, and POI name for destination input. Register in the voice dictionary 54.
[0018]
As shown in FIG. 8, the destination input speech dictionary 54 has a speech dictionary portion 54a used when inputting a destination address / intersection and a speech dictionary portion 54b used when inputting a POI name. In the voice dictionary unit 54a, a state name list 61, a city name list 62, and a street name list 63 are hierarchically formed. The state name list 61 stores the correspondence between the state name (character string), the voice pattern, and the pointer that points to the list of city names belonging to the state. The state name list 61 is preset because it has a small number of states and is important. Whenever a user specifies a state when inputting a destination, the state name list 61 synthesizes and stores a sound pattern of the state name. You may make by. The city name list 62 is provided for each state. When a city name is keyed in when entering a destination, a correspondence between a city name (character string), a voice pattern, and a pointer indicating a street name list belonging to the city is created in the city name list of the state to which the city belongs. be registered. The street name list 63 is provided for each city. When a street name is key-input at the destination input, a correspondence between the street name (character string) and the voice pattern is created and registered in the street name list of the city to which the street belongs.
The voice dictionary unit 54b stores the POI name and the voice pattern input when the destination is input by the POI name.
[0019]
(C) Destination Input Speech Dictionary Creation Processing FIG. 9 is a flowchart of destination entry speech dictionary creation processing.
When inputting the destination, the navigation device 10 determines whether the destination is input by an address or an intersection or a POI name (step 101). If the POI name is input, the POI name is keyed. If the POI name is input, the data for specifying the destination input method and the character string of the POI name are input to the speech recognition apparatus (step 103). The speech dictionary creation unit 56b of the speech recognition device checks whether a POI name speech pattern has already been registered (step 104). If not, the POI name speech pattern is synthesized (step 105). The correspondence of the voice pattern is registered in the voice dictionary section 54b of the destination input voice dictionary 54 (step 106).
[0020]
In step 101, when the destination is input by an address or an intersection, it is monitored whether the city name is key-input (step 111). If the city name is input, the data for specifying the destination input method and the city name are checked. The character string is input to the voice recognition device (step 112). The speech dictionary creation unit 56b of the speech recognition device checks whether or not a city name speech pattern has already been registered (step 113). If not registered, the city name speech pattern is synthesized (step 114). The correspondence of the voice pattern is registered in the voice dictionary unit 54a of the destination input voice dictionary 54 (step 115).
Next, it is monitored whether the street name has been input (step 121). If the street name is input, a character string of the street name is input to the voice recognition device (step 122). The speech dictionary creation unit 56b of the speech recognition apparatus checks whether a street name speech pattern has already been registered (step 123). If not, the street name speech pattern is synthesized (step 124). The correspondence of the voice pattern is registered in the voice dictionary section 54a of the destination input voice dictionary 54 (step 125).
[0021]
(D) Destination Speech Recognition Processing FIG. 10 shows a destination speech recognition processing flow. It is assumed that the navigation apparatus has already notified the voice recognition apparatus by determining whether the destination is input by an address or an intersection or a POI name based on the user selection. When there is a voice input in such a state (step 201), the voice recognition unit 56a of the voice recognition apparatus determines whether the destination is input by an address or an intersection or a POI name (step 202), and the POI name If it is an input, the POI name having the voice pattern most similar to the input voice is searched with reference to the voice dictionary unit 54b, and the POI name is input to the navigation device (step 203).
[0022]
On the other hand, in step 202, if the destination is input by address or intersection, it is checked whether the current state is the state name input state. If the destination is the state name input state, refer to the state name list 61 of the speech dictionary unit 54a. The state name having the voice pattern most similar to the input voice is searched, and the state name is input to the navigation device (steps 204 to 205). If it is not the state name input state but the city name input state, the city name list 62 corresponding to the state where the voice is input is referred to search for the city name having the voice pattern most similar to the input voice, and the city name is Input to the navigation device (steps 206 to 207). If the street name input state is not the state name or city name input state, the street name list 63 corresponding to the voice input city is referred to search for the street name having the voice pattern most similar to the input voice, The street name is input to the navigation device (step 208).
In the above, the case of destination input in the United States has been described, but the present invention can be applied to other countries as it is or with some corrections in addition to the United States.
The present invention has been described with reference to the embodiments. However, the present invention can be variously modified in accordance with the gist of the present invention described in the claims, and the present invention does not exclude these.
[0023]
【The invention's effect】
As described above, according to the present invention, only the voice pattern of the area name and facility name corresponding to the destination required by the user needs to be added and registered to the initial registered voice pattern. It can be reduced and the destination can be correctly recognized with less hardware.
In addition, according to the present invention, when inputting a destination by gradually narrowing from an upper area of a large area to a lower area of a small area, the area of the lower area belonging to the upper area corresponding to the upper area When a name and its synthesized speech are registered, and the upper input region is specified, the lower region is input by referring to the region name of the lower region belonging to the upper region and its synthesized speech in the next speech input Therefore, the range for searching for a speech pattern similar to the input speech is narrowed, and speech recognition can be performed with high accuracy in a short time.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a navigation device of the present invention.
FIG. 2 is an explanatory diagram (part 1) of a method for specifying a destination by inputting an address;
FIG. 3 is an explanatory diagram (part 2) of a destination specifying method based on address input;
FIG. 4 is a keyboard image.
FIG. 5 is an explanatory diagram of a destination specifying method by inputting a POI name.
FIG. 6 is a block diagram of a speech recognition apparatus according to the present invention.
FIG. 7 is an explanatory diagram of phonemes.
FIG. 8 is an explanatory diagram of a destination input speech dictionary;
FIG. 9 is a flowchart of a destination input speech dictionary creation process.
FIG. 10 is a voice recognition processing flow of a destination.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 Navigation apparatus 51 Microphone 52 Speech dictionary database 53 Phoneme pattern database 54 Destination input speech dictionary 55 Communication interface unit 56 Speech recognition engine 56a Speech recognition unit 56b Speech dictionary creation unit

Claims (4)

In a destination input method of a navigation device for inputting a destination by key input and voice input,
When the name of the area to which the destination belongs or the name of the facility as the destination is keyed in the destination input mode, a process for specifying the destination is performed using the name of the area or the name of the facility entered by the key . Synthesizing the voice of the area name or facility name input by the key , and registering the synthesized voice in the voice dictionary corresponding to the area name or facility name;
In the destination input mode, when the name of the area to which the destination belongs or the name of the facility as the destination is inputted by voice, the name of the area to which the inputted destination belongs or the name of the facility as the destination is identified by referring to the voice dictionary And
Using the name of the area to which the identified destination belongs or the name of the facility as the destination, to perform processing for specifying the destination;
A destination input method by voice recognition. When inputting a destination by gradually narrowing from a higher area of a large area to a lower area of a small area, register the area name of the lower area belonging to the upper area and its synthesized speech in association with the upper area. Leave
When the upper region input by voice is specified, in the next voice input, the region name of the lower region belonging to the upper region and its synthesized voice are referred to recognize the lower region input by voice.
The destination input method according to claim 1, wherein: Decompose keynamed region name or facility name into phonemes and synthesize phoneme patterns according to phonemes to synthesize sounds of region names or facility names.
The destination input method according to claim 1, wherein: In a navigation system for inputting a destination by key input and voice input,
Key input means for keying in the name of the area to which the destination belongs or the name of the facility as the destination in the destination input mode;
Speech synthesis unit to synthesize the voice of the key input area name or facility name,
Voice dictionary unit to register a voice that was the synthesis so as to correspond to the local name or facility name,
When the facility name as a regional name or destination belongs destination has been voice input, said with reference to the voice dictionary unit the voice input destination to identify the facility name as a regional name or destination belong voice Recognition part,
In the destination input mode, a navigation control unit that performs processing for specifying a destination using the name of a region to which the destination input by key or voice belongs or the name of a facility as a destination,
A navigation system characterized by comprising:

Images