JP6187366B2 - Image processing apparatus, image processing method, and program - Google Patents

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program.

  2. Description of the Related Art Conventionally, there is an image processing apparatus that extracts and stores image data of a specific area (article) from image data obtained by reading a document such as a newspaper or a magazine.

  In this type of image processing apparatus, marking of a specific area is necessary for extracting image data of the specific area. That is, in order to acquire image data of a specific area, a user marks a specific area in a document with a pen (marker) or the like, and then causes the scanner attached to the image processing apparatus to read the marked document. In the image processing apparatus, marking is detected from the image data of the read document, a specific area is specified from the marking, and image data of the specific area is extracted.

JP 2002-150306 A

  However, since marking of a specific area is required, there is a problem that it takes time for marking work. In addition, the document is soiled by the marking.

  In addition, if the extracted image data of the specific area is stored as it is, the image is output in the layout on the original when the data relating to the image data is output on the display screen or the sheet. Is difficult to read. Patent Document 1 proposes that image data in a specific area is laid out in a rectangular area and the laid out image data is output. However, articles such as newspapers are arranged in columns, and if the image data is laid out in a rectangular area ignoring the columns, the output image may be difficult to read.

  An object of the present invention is to synthesize image data so that marking for extracting image data of a specific article can be made unnecessary, and an image related to the extracted image data can be easily viewed by a user. An image processing apparatus, an image processing method, and a program are provided.

  In order to achieve the above object, an image processing apparatus according to the present invention includes an image acquisition unit that acquires image data, and a column separator that separates an image of a source document into predetermined column widths for each source document. A storage unit for storing information and the column width between the column separators in association with each other, and a control unit, wherein the control unit searches for column separators based on information on the column separators. A separator search process, a search process for searching for a heading image and an end image from the image data acquired by the image acquisition unit, and a column separator from the heading image to the end image detected by the search process. For each column area of the column width, there is an extraction process for extracting image data and a synthesis process for combining the image data extracted for each column area by the extraction process.

  According to this configuration, the column separator is searched from the image data, and the headline image and the end image of the article are searched. When a heading image and an end image are detected by this search, image data from the heading image to the end image is extracted for each column region, that is, for each region between column separators adjacent to each other in the column width direction. The Therefore, marking for extracting image data of an article can be made unnecessary.

  Then, the image data extracted for each column area is synthesized. Therefore, it is possible to obtain image data synthesized taking column sets into consideration. As a result, it is possible to make the image related to the combined image data easy to see for the user.

  The control unit may further include an input request process for requesting input of source data and heading information of image data.

  According to this configuration, by inputting the source document and heading information of the image data, it becomes easy to detect the heading image and the end image, and the image data desired by the user can be extracted with high accuracy.

  The search process includes, for each column area, a header image search process for searching for a header image, and a column area including the header image detected by the header image search process in a predetermined direction along the column separator from the header image. If the end image is searched and no end image is detected, the next column area adjacent to the column area in the direction orthogonal to the predetermined direction is searched for and the end image is searched for in the predetermined direction. In addition, it may include an end image search process for executing a search target movement and an end image search until the end image is detected.

  According to this configuration, a heading image is searched for each column region. When a heading image is detected, an end image is searched for a column area including the heading image in a predetermined direction from the heading image along the column separator. If the end image is not detected in the column area, the column area is successively searched for the end image until the end image is detected. Thereby, the headline image and the end image can be detected efficiently and accurately.

  It is preferable that the control unit searches for a region having a width shorter than the column width from the column separator in the header image search process.

  According to this configuration, the heading image can be detected in a short time because the region in which the heading image is searched is narrower than the configuration in which the heading image is searched in the entire column area.

  For each source document, the storage unit stores information on an article separator that divides the image of the source document in the predetermined direction, and the extraction process is image data of a column area including the header image, and the header image A first extraction process for extracting image data from a headline image to an article separator located downstream in a predetermined direction, and image data of a column area other than the column area including the header image, A second extraction process for extracting image data from an article separator on the upstream side in the direction to an article separator on the downstream side or the end image, and the control unit includes the image data extracted by the first extraction process in the synthesis process, The image data extracted by the second extraction process may be combined with the article separator as a boundary.

  According to this configuration, the image data from the headline image to the article separator on the downstream side in the predetermined direction in the column area including the headline image, and the article separator on the downstream side in the predetermined direction to the article separator or end image in the predetermined direction. Are extracted. Then, the extracted plurality of image data are combined with the article separator as a boundary. Therefore, an image related to the synthesized image data can be obtained as one image that is easy for the user to see.

  In the synthesis process, the control unit relates to the synthesized image data when the image data extracted by the first extraction process and the image data extracted by the second extraction process are synthesized using the article separator as a boundary. If the width of the image in the predetermined main scanning direction exceeds the specified width, the image data extracted by the first extraction process and the image data extracted by the second extraction process are connected to the upstream article in the predetermined direction. You may synthesize | combine by aligning with a separator.

  According to this configuration, when a plurality of image data are combined with the article separator as a boundary, for example, when the combined image data is output to a sheet by an image forming apparatus such as a printer, It is assumed that the width of the output image exceeds the specified width that can be output by the image forming apparatus in the scanning direction. In this case, the plurality of pieces of image data are combined and aligned with the article separator on the upstream side in the predetermined direction, whereby an image related to the combined image data can be obtained as one image that is easy for the user to view.

  When the end image cannot be detected from the image data acquired by the image acquisition unit in the end image search process, the control unit searches for the end image from the image data acquired next by the image acquisition unit. May be.

  According to this configuration, the end image can be detected even if the end image is included in the image data acquired next by the image acquisition unit. As a result, even when the image data desired by the user extends over a plurality of pages, a plurality of image data on different pages can be combined into one image data.

  The control unit further detects the direction of characters included in the image related to the image data acquired by the image acquisition unit, and applies to the image data so that the direction of the characters matches the direction of the source document. It is preferable to have a rotation process for executing a process of rotating the image, and in the search process, search for the heading image and the end image from the image data after the rotation process.

  According to this configuration, the direction of characters included in the image related to the image data acquired by the image acquisition unit is detected, and the image data is imaged so that the direction of the characters matches the direction of the source document. Is rotated. Since the heading image and the end image are searched from the image data after the rotation processing, the heading image and the end image can be detected with high accuracy. As a result, the image data desired by the user can be extracted satisfactorily.

  Note that the present invention can be realized not only in the form of an image processing apparatus, but also in the form of a program for realizing, for example, an image processing method and a computer function of the image processing apparatus.

  According to the present invention, in the image processing apparatus, it is possible to eliminate the need for marking for extracting image data of a specific article, and to make it easy for the user to view the image related to the extracted image data. Data can be synthesized.

FIG. 2 is a perspective view of the MFP according to the embodiment of the present invention as viewed from the upper left front side. 2 is a block diagram illustrating an electrical configuration of the MFP. FIG. It is a figure which shows an example of a marker table. It is a figure which shows an example of a source information table. It is a flowchart which shows the flow of an article extraction process. FIG. 3 is a conceptual diagram illustrating a reading state of a document image. FIG. 6 is a conceptual diagram for explaining image rotation processing of a read document image. It is a flowchart which shows the flow of a heading marker search process. It is a flowchart which shows the flow of an end marker search process. It is a conceptual diagram which shows an example of an article. It is a conceptual diagram for demonstrating the search of a column separator and a heading marker. It is a conceptual diagram for demonstrating the search of an end marker. It is a conceptual diagram for demonstrating the case where an end marker is not detected. It is a conceptual diagram for demonstrating the search of an end marker when an end marker is not detected by the last scan. It is a flowchart which shows the flow of an extraction synthetic | combination process. It is a conceptual diagram for demonstrating an example of an extraction synthetic | combination process. It is a conceptual diagram for demonstrating the other example of an extraction synthetic | combination process.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

<Mechanical configuration>

  An MFP (Multi-Function Peripheral) 1 as an example of an image processing apparatus includes a printer main body 2, a scanner main body 3, and an ADF (Auto Document Feeder) 4, as shown in FIG.

  In the following description, the up / down, left / right, and front / rear directions defined when the MFP 1 placed on a plane is viewed from the side where an operation panel 7 (described later) is provided are used.

  An image forming unit 10 (see FIG. 2) is accommodated in the printer main body 2. The image forming unit 10 performs printing (color printing or monochrome printing) based on image data, and forms an image on a sheet. The image forming method may be an electrophotographic method or an inkjet method.

  A contact glass 8 is provided on the upper surface of the scanner body 3. The scanner body 3 houses an image reading unit 20 (see FIG. 2). The image reading unit 20 includes a CIS (Contact Image Sensor) extending in the front-rear direction. The image reading unit 20 as an example of the image acquisition unit moves the CIS 9 (see FIG. 5A) to the right side, reads an image formed on the sheet placed on the contact glass 8 from the sheet, and outputs it as image data. It is stored in the RAM 73.

  A paper feed tray 5 is provided at the bottom of the printer body 2. The paper feed tray 5 is configured to accommodate a plurality of sheets in a stacked state. A paper discharge tray 6 is formed on the upper surface of the printer main body 2. The sheets accommodated in the paper feed tray 5 are sent one by one from the paper feed tray 5 into the printer main body 2 and conveyed through the printer main body 2 toward the paper discharge tray 6. While the sheet is conveyed through the printer main body 2, an image is formed on the sheet by the image forming unit 10. The sheet (printed material) on which the image is formed is discharged to the paper discharge tray 6.

  The scanner body 3 is disposed above the printer body 2, and the ADF 4 is disposed above the scanner body 3.

  The MFP 1 also includes an operation panel 7. The operation panel 7 is disposed, for example, at the front end of the scanner body 3. The operation panel 7 is provided with a display unit 30 and an operation unit 40.

  The display part 30 consists of a liquid crystal display, for example. Various information is displayed on the display unit 30.

  The operation unit 40 includes various operation buttons. Various inputs are possible by pressing the operation buttons. For the operation buttons, for example, an OK button for confirming various instructions (for example, an instruction to start image processing), a cancel button for canceling various instructions, and a key displayed on the display unit 30 are selected. A cross button, numeric keypad for entering numbers and characters, etc. are included.

<Electrical configuration>

  As shown in FIG. 2, the MFP 1 includes a communication interface (I / F) 60 and a control unit 70.

  A communication interface 60 as an example of a reception unit is an interface for communication with an external device such as a PC (personal computer) connected to a LAN (Local Area Network). The communication method may be a wireless communication method or a wired communication method.

  The control unit 70 includes an ASIC (Application Specific Integrated Circuit) 71, a ROM 72, a RAM 73, and an NVRAM 74.

  The ASIC 71 incorporates a CPU 75. An instruction by the operation of the operation unit 40 is input to the ASIC 71. In addition, a print job transmitted from an external device is input to the ASIC 71 via the communication interface 60. FIG. 2 shows a PC (Personal Computer) 80 as an example of an external device. The CPU 75 controls the image forming unit 10 and the display unit 30 by executing programs for various processes based on information input to the ASIC 71 and controls communication via the communication interface 60.

  In the following, a case where a print job transmitted from the PC 80 is accepted by the MFP 1 is taken as an example.

  The ROM 72 stores programs executed by the CPU 75 and various data.

  The RAM 73 and NVRAM 74 are used as work areas when the CPU 75 executes the program.

<Marker table and source information table>

  The NVRAM 74 as an example of the storage unit stores a marker table shown in FIG. 3A and a source information table shown in FIG. 3B.

  In the marker table, the name, source, heading marker, and end marker of the article to be extracted are stored in association with each other.

  The name is a name given to identify the article.

  The source is information indicating the source of the article, and is, for example, a newspaper name or a magazine name.

  The headline marker is a marker that indicates the headline of the article, and is, for example, a symbol that indicates the headline in the article or a name that is used for the headline in the article.

  The end marker is a marker indicating the end of the article, and is, for example, a symbol indicating the end of the article or a character string indicating the end of the article.

  In the marker table shown in FIG. 3A, in the first line, “serialized article” as name, “XX newspaper” as source, heading symbol as heading marker, and end symbol as end marker are stored in association with each other. . In the second line, the name “weather forecast”, the source “◯ △ newspaper”, the heading marker “weather forecast”, and the absence of the end marker are stored in association with each other.

  In the source information table, the name of the article to be extracted, the source, the character direction, the column width, the column separator, and the article separator are stored in association with each other.

  The name is a name given to identify the article.

  The source is information indicating the source of the article, and is, for example, a newspaper name or a magazine name.

  The character direction is information indicating whether the article is written vertically or horizontally.

  The column width is information indicating the column width of the article.

  A column separator is a symbol indicating a space between columns in the same article.

  The article separator is a pattern indicating between an article and another article.

  In the source information table shown in FIG. 3B, in the first line, the name “serialized article”, the source “XX newspaper”, the character direction “vertical”, the column width “3 cm”, and the column separator The article separator design in the article is stored in association with the column separator design and article separator in the article. In the second line, the name is “weather forecast”, the source is “○ △ newspaper”, the text direction is “vertical”, the column width is “2.5 cm”, there is no column separator, there is no article separator Are stored in association with each other.

<Article extraction process>

  While the MFP 1 is powered on, the CPU 75 of the ASIC 71 repeatedly executes the article extraction process shown in FIG.

  In the article extraction process, the CPU 75 executes a scan process of a document (source document) in which an article to be extracted is described (S1). Execution of the scanning process means that the image reading unit 20 is controlled by the CPU 75, the image of the original is read by the image reading unit 20, and image data based on the image is stored in the NVRAM 74. As shown in FIG. 5A, a document is placed on the contact glass 8 of the MFP 1 and an image of the surface of the document facing the contact glass 8 is read.

  The source manuscript is, for example, an article such as a newspaper or a magazine.

  As shown in FIG. 8, newspaper articles are vertically written and vertically arranged. The upper and lower boundaries of the article and each level are separated by a column separator. Hereinafter, an area between column separators is referred to as a “column area”. The left and right boundaries of the article are separated by an article separator. Further, in many cases, there is a heading (heading marker) at the top of each article (the right end at the top of the article), and an end marker at the end of each article (the left end at the bottom of the article).

  For example, for major articles of famous newspapers and magazines, the heading marker and end marker are registered in advance in the marker table, and the character direction, column width, column separator and article separator are registered in advance in the source information table. Yes. Hereinafter, the heading marker, end marker, character direction, column width, column separator, and article separator are referred to as “article specifying information”.

  In the following, the flow of processing will be described using a newspaper article as an example.

  The CPU 75 executes an image rotation process on the scanned image (S2). As shown in FIG. 5A, the document is placed inclined with respect to the main scanning direction that is the longitudinal direction of the CIS 9 and the sub-scanning direction that is orthogonal to the CIS 9 and extends along the surface of the contact glass 8. There is. As shown in FIG. 5B, the image rotation process recognizes the orientation of characters in the read image, and converts, for example, conversion into image data so that the image is rotated so that the upper side of the character is upstream in the sub-operation direction. Against.

  Next, the CPU 75 determines whether or not an end marker has been detected in the previous scan (S3).

  If an end marker is detected in the previous scan (S3: YES), this scan is to extract a new article.

  The user can register a new heading marker and end marker that are not currently registered in the marker table in the marker table. Hereinafter, the heading marker and the end marker are referred to as “marker information”. Registration of new marker information in the marker table is performed by causing the MFP 1 to transmit the marker information and the name and source of the article in which the marker information is used from the PC 80 connected to the MFP 1 via the communication interface 60. .

  New marker information transmitted from the PC 80 to the MFP 1 is temporarily stored in the RAM 73 of the MFP 1. Thereafter, the CPU 75 causes the display unit 30 to display a selection screen as to whether or not to register the received new marker information in the marker table in the NVRAM 74, and whether the selection to register the new marker information has been input by the operation unit 40. It is determined whether or not (S4).

  When selection for registering new marker information is input (S4: YES), the CPU 75 causes the operation unit 40 to input marker information by causing the display unit 30 to display a new marker information registration screen. The inputted information is registered in the marker table as new marker information (S5).

  When selection not registering new marker information is input (S4: NO), or after registration processing of new marker information is executed (S5), the CPU 75 displays an input request for article selection information on the display unit 30. (S6) until the input of article selection information by the operation unit 40 is completed (S7: NO), the article selection information is input (S7). The article selection information is, for example, an article name and a source. In the display of the input request for the article selection information, for example, the name of the article to be extracted and an image requesting input of the source are displayed on the display unit 30, and the name and source of the article are input by the operation unit 40. Thereby, the article specifying information used for extraction of the extraction target article can be acquired from the marker table and the source information table.

  On the other hand, when the end marker is not detected in the previous scan (S3: NO), the CPU 75 determines whether or not to link the image data extracted in the previous scan to the image data extracted in the current scan. A selection screen is displayed on the display unit 30, and it is determined whether or not a selection to be connected has been input by the operation unit 40 (S8).

  When the selection not to be connected is input (S8: NO), the extraction of a new article is started, so the CPU 75 is the same as when the end marker is detected in the previous scan (SS3: YES). Then, a screen for selecting whether or not to register new marker information is displayed on the display unit 30, and it is determined whether or not a selection for registering new marker information has been input by the operation unit 40 (S4).

  When the selection to connect is input (S8: YES), the continuation of the article extracted in the previous scan is extracted. Therefore, the CPU 75 does not perform the processes of steps S4 to S7 (skip of S4 to 7) and executes the heading marker search process (S9).

  On the other hand, even when the input of the article identification information is completed (S7: YES), the CPU 75 executes a heading marker search process (S9). The heading marker search process is a process of detecting the article extraction start position for the read image data. Details of the heading marker search process will be described later.

  Then, the CPU 75 executes an end marker search process (S10). The end marker search process detects the end image of article extraction from the read image data, and extracts each column area existing between the start position and the end image position as image data for each column area. , Processing stored in the RAM 73. Details of the end marker search process will be described later.

  Next, the CPU 75 causes the display unit 30 to display a selection screen as to whether or not there is a scan for the next page of the document, and whether or not a selection indicating that there is a scan for the next page has been input by the operation unit 40. Judgment is made (S11).

  If there is a scan for the next page (S11: YES), the process returns to step S1, and the CPU 75 executes a scan process for the next page (S1).

  If there is no scan for the next page (S11: NO), the CPU 75 executes an extraction / synthesis process (S12) and ends the article extraction process. The extraction / combination processing is processing that combines the image data of articles for each columnar area extracted so far and connects the articles in one image data. Details of the extraction and synthesis process will be described later.

<Header marker search process>

  The flow of the header marker search process executed in step S9 of FIG. 4 is shown in FIG.

  In the headline marker search process, the CPU 75 determines whether or not a temporary headline marker is stored in the RAM 73 (S901). The temporary headline marker is the last search area stored in the RAM 73 as a temporary headline marker when the end marker is not detected in the previous search for the end marker.

  When the temporary headline marker is not stored in the RAM 73 (S901: NO), the headline marker is searched for a new article. Therefore, the CPU 75 stores the article specifying information (heading marker, end marker, character direction, column width, column separator, and article separator) associated with the name and source input in step S7 of FIG. Obtained from the source information table (S902).

  Then, the CPU 75 starts searching for the column separator (S903), and searches for the column separator (S904) until the column separator is detected (S904: NO).

  In the column separator search, as shown in FIG. 9A, for example, an image of a predetermined size, for example, a 5 mm square is moved at a predetermined interval, for example, 5 mm while moving the right end in the article area of the document from the upper end to the lower side. When extracted at each interval, it is determined whether or not a part of the extracted image matches a part of the image of the column separator (column separator obtained from the source information table) of the article to be extracted. Assume that a column separator is detected. The article area is an area excluding the margin when there are margins (blank portions) at the top, bottom, left, and right edges of the document.

  If a column separator is detected (S904: YES), the CPU 75 searches for a heading marker (S905) and determines whether a heading marker is detected (S906).

  As shown in FIG. 9A, the search for the heading marker first places a search point at the upper right end in the lower column area of the detected column separator and extracts an image of the search point. The search point is, for example, a predetermined width (for example, 3 mm), and a column separator and a predetermined distance downward from the column separator (for example, a distance of one third of the column width of the article to be extracted). With the area between. Then, it is determined whether or not a part of the extracted image matches a part of the heading marker image of the article to be extracted, and the heading marker is detected when they match.

  When the heading marker is detected (S906: YES), the CPU 75 stores the position of the detected heading marker on the original image in the RAM 73 (S907), and ends the heading marker search process.

  When the heading marker is not detected (S906: NO), the CPU 75 moves the search point of the heading marker on the document image to the left by a predetermined distance, for example, 5 mm, and acquires the image data at that position ( S908).

  Then, the CPU 75 determines whether it is an article separator and whether the search point is the left end of the article area (S909).

  Whether or not the search point is the article separator is determined by, for example, whether or not a part of the extracted image matches a part of the article separator image of the extraction target article. Whether or not the search point is the left end of the article area is determined from the size of the article area and the position of the search point, for example.

  If the search point is the article separator or the left end of the article area (S909: YES), the CPU 75 moves the search point to the right end of the next (lower) column area (S910). Returning to step S905, the retrieval of the heading marker is executed again for the column area (S905).

  When the search point is not the article separator and is not the left end of the article area (S909: NO), the CPU 75 compares the image of the moved search point area with the heading marker image of the article to be extracted. It performs (S905).

  On the other hand, when the temporary headline marker is stored in the RAM 73 (S901: YES), since the image extraction is completed in the middle of the article in the previous scan, the uppermost column area of the article area is not detected. The right end is stored in the RAM 73 as the position of the heading marker (S911), and the CPU 75 ends the heading marker search process.

<End marker search process>

  The flow of the end marker search process executed in step S10 of FIG. 4 is shown in FIG.

  In the end marker search process, the position of the heading marker stored in the RAM 73 in step S907 of FIG. 6 or step S911 is set as the search start point (S1001).

  Then, the CPU 75 searches for an end marker (S1002), and determines whether an end marker has been detected (S1003).

  In the search for the end marker, as shown in FIG. 9B, first, an image of the search area at the search start point (for example, an area between the upper and lower column separators with a width of 10 mm) is extracted. Then, it is determined whether or not a part of the extracted image matches a part of the image of the end marker of the extraction target article, and if it matches, the end marker is detected.

  When the end marker is not detected (S1003: NO), the CPU 75 moves the search area to the left by a predetermined distance, for example, 15 mm (S1004).

  Then, the CPU 75 determines whether the search area is an article separator or the left end of the article area (S1005).

  Whether or not the search area is an article separator is determined by whether or not a part of the image in the search area matches a part of the article separator image of the extraction target article. Whether or not the search area is the left end of the article area is determined from the size of the article area and the position of the search area.

  When the search area is the left end of the article area (S1005: YES), the CPU 75 stores the image data of the current column area from the search start point to the left end of the article area in the RAM 73 (S1006). Is the article separator (S1005: YES), the CPU 75 stores the image data of the current column area from the search start point to the article separator in the RAM 73 (S1006).

  Then, the CPU 75 determines whether or not the search area can be moved to the next (lower) column area (S1007).

  When the search area cannot be moved to the next column area, as shown in FIG. 10A, for example, there is no next column area on the current page.

  If it is possible to move the search area to the next column area (S1007: YES), the CPU 75 moves the search area to the right end of the next column area (S1008), and the process of step S1002 Returning to step S1002, the end marker is searched for.

  If it is not possible to move the search area to the next column area (S1007: NO), as shown in FIG. 10A, it means that the search for the current page has been completed. The last search area in FIG. 6 is stored in the RAM 73 as a temporary headline marker (S1009), and, as shown in FIG. 10B, the process ends in step S911 in FIG. 6 with the right edge of the next page as the search start point. Perform marker search processing.

  On the other hand, when the end marker is detected (S1003: YES), the CPU 75 stores the image data of the current column area from the search start point to the end marker in the RAM 73 (S1010), and performs end marker search processing. finish.

<Extract synthesis process>

  FIG. 11 shows the flow of the extraction and synthesis process executed in step S11 of FIG.

  In the extracting and synthesizing process, the CPU 75 determines whether there is image data from the heading marker to the article separator in the image data stored in the RAM 73 for this article (S1101).

  If there is no image from the heading marker to the article separator (S1101: NO), it means that the article is completed with only one column area, so the image data stored in the RAM 73 is used as the article to be extracted. decide. At this time, for example, an image may be formed on the sheet for this image data.

  If there is an image from the heading marker to the article separator (S1101: YES), the CPU 75 extracts image data from the search start point to the article separator among the plurality of image data stored in the RAM 73 for this article. (S1102). The image data from the search start point to the article separator is image data that does not include an end marker such as image data from the heading marker to the article separator and image data from the article separator to the article separator.

  Next, the CPU 75 extracts image data from the search start point to the end marker among the plurality of image data stored in the RAM 73 for this article (S1103). The image data from the search start point to the end marker is image data that does not include the heading marker but includes the end marker, like the image data from the article separator to the end marker.

  Then, as shown in FIG. 12A, the CPU 75 extracts the image data extracted from the RAM 73 for this article in the order stored in the RAM 73, and the position of the article separator on the left side of the article in the image data and the article in the next image data. The image data is synthesized so as to match the position of the right article separator (S1104).

  At this time, for example, when the synthesized image data is output to the sheet by the MFP 1, if the width of the output image exceeds the specified width that can be output by the MFP 1 in the main scanning direction of the MFP 1, FIG. As shown, each image data is aligned with the article separator on the right side of the article and synthesized.

  Then, the CPU 75 stores the synthesized image data in the RAM 73 as an extraction target article (S1105). At this time, for example, an image may be formed on the sheet for this image data.

<Effect>

  As described above, the headline marker and the end marker of the article are searched from the image data. When a heading marker and an end marker are detected by this search, image data from the heading marker to the end marker is extracted for each column region, that is, for each region between column separators adjacent to each other in the column width direction. The Therefore, marking for extracting image data of an article can be made unnecessary.

  Then, the image data extracted for each column area is synthesized. Therefore, it is possible to obtain image data synthesized taking column sets into consideration. As a result, it is possible to make the image related to the combined image data easy to see for the user.

  The CPU 75 requests input of article selection information for specifying article specifying information and marker information of the extraction target article.

  By inputting the article selection information, the article specifying information and the marker information of the extraction target article are specified, the detection of the heading image and the end image is facilitated, and the image data desired by the user can be extracted with high accuracy. it can.

  The CPU 75 searches for a heading marker for each column area. If the end marker is searched for the column area including the detected headline marker in the direction of the sentence flow along the column separator from the headline marker, and the end marker is not detected, the column area The next column area adjacent to is searched for the end marker in the direction of the sentence flow, and the search target is moved and the end marker is searched until the end marker is detected.

  Thereby, a heading marker is searched for every column area. When the heading marker is detected, the end marker is searched for the column region including the heading marker in the direction of the sentence flow along the column separator from the heading marker. When the end marker is not detected in the column area, the column area is successively searched for the end marker until the end marker is detected. Thereby, a heading marker and an end marker can be detected efficiently and accurately.

  The CPU 75 searches for an area having a width shorter than the column width from the column separator.

  As a result, the heading marker can be detected in a short time because the area where the heading marker is searched is narrower than the configuration in which the heading marker is searched in the entire column area.

  The NVRAM 74 stores, as a source information table, information on article separators that classify images of source documents in the direction of the flow for each source document. The CPU 75 extracts image data from the column marker area including the heading marker to the article separator located downstream of the sentence with respect to the heading marker, and includes the column marker including the heading marker. Image data from a column area other than the area, from the article separator on the upstream side of the sentence to the article separator or end marker on the downstream side is extracted. Then, when there are a plurality of extracted image data, the CPU 75 combines the extracted image data with the article separator as a boundary.

  As a result, image data from the headline marker to the article separator on the downstream side of the sentence in the column area including the headline marker, and image data from the article separator on the upstream side of the sentence to the article separator or end marker on the downstream side of the sentence. And are extracted. Then, the extracted plurality of image data are combined with the article separator as a boundary. Therefore, an image related to the synthesized image data can be obtained as one image that is easy for the user to see.

  When the extracted image data is combined with the article separator as a boundary, the CPU 75 extracts the extracted image data if the width of the sentence flow direction in the image related to the combined image data exceeds the specified width. A plurality of image data are combined and aligned with the article separator on the upstream side of the sentence.

  As a result, when a plurality of image data are combined with the article separator as a boundary, for example, when the combined image data is output to the sheet by the MFP 1, the width of the output image in the main scanning direction of the MFP 1 is MFP1. It is assumed that the output exceeds the specified width that can be output with. In this case, by combining a plurality of image data in alignment with the article separator on the upstream side of the sentence, it is possible to obtain an image related to the combined image data as one image that is easy for the user to view.

  If the end marker cannot be detected from the scanned image data, the CPU 75 searches for the end marker from the next scanned image data.

  Thereby, even if the end marker is included in the image data scanned next time, the end marker can be detected. As a result, even when the image data desired by the user extends over a plurality of pages, a plurality of image data on different pages can be combined into one image data.

  The CPU 75 detects the direction of characters included in the image related to the scanned image data, and executes a process of rotating the image with respect to the image data so that the direction of the character matches the direction of the source document. To do. Then, the heading marker and the end marker are searched from the image data after the rotation processing.

  As a result, the direction of characters included in the image related to the scanned image data is detected, and processing for rotating the image with respect to the image data is performed so that the direction of the character matches the direction of the source document. . Since the heading marker and the end marker are searched from the image data after the rotation processing, the heading marker and the end marker can be detected with high accuracy. As a result, the image data desired by the user can be extracted satisfactorily.

<Modification>

  As mentioned above, although embodiment of this invention was described, this invention can also be implemented with another form.

  For example, in the above-described embodiment, a vertically written newspaper article has been described as an example of a source manuscript, but the source manuscript may be a horizontally written article. In the case of a horizontally written article, it is arranged in the horizontal direction of the document, so the search for the column separator is performed from the upper end of the column area in the right column again to the left. In this case, the next column area means a column area on the left side of the current column area.

  In addition, various design changes can be made to the above-described configuration within the scope of the matters described in the claims.

1 MFP
20 Image reading unit 70 Control unit 74 NVRAM

Claims (10)

An image acquisition unit for acquiring image data;
For each source document, a storage unit that stores information on column separators that classify images of source documents into predetermined column widths and the column widths between column separators are stored in association with each other.
A control unit;
With
The controller is
A column separator search process for searching for the column separator based on the information of the column separator;
A search process for searching for a heading image and an end image from the image data acquired by the image acquisition unit;
From the heading image detected in the search process to the end image, an extraction process for extracting image data for each column area of the column width along the column separator;
An image processing apparatus comprising: a composition process for composing each image data extracted for each column area by the extraction process. The image processing apparatus according to claim 1,
The controller is
Furthermore,
An input request process for requesting input of source image information and heading information of the image data;
An image processing apparatus. The image processing apparatus according to claim 1, wherein:
The search process includes
For each column region, a heading image search process for searching the heading image;
When the end image is searched in the predetermined direction along the column separator from the heading image for the column region including the heading image detected by the heading image search process, and the end image is not detected Search for the next column area adjacent to the column area in a direction orthogonal to the predetermined direction until the end image is searched in the predetermined direction until the end image is detected. An end image search process for executing the movement of the image and the search for the end image;
An image processing apparatus. The image processing apparatus according to claim 3,
The controller is
In the heading image search process, a region having a width shorter than the column width is searched from the column separator.
Image processing device. The image processing apparatus according to claim 3 or 4,
In the storage unit, for each source document, information of an article separator that classifies the image of the source document in the predetermined direction is stored.
The extraction process is:
A first extraction process for extracting image data from the header image to an article separator located downstream in the predetermined direction with respect to the header image, the image data of the multi-column area including the header image; ,
A second extraction that extracts image data of a column area other than the column area including the heading image, from the upstream article separator to the downstream article separator or the end image in the predetermined direction. Processing,
Including
The controller is
In the synthesis process, the image data extracted by the first extraction process and the image data extracted by the second extraction process are synthesized using an article separator as a boundary.
Image processing device. The image processing apparatus according to claim 5,
The controller is
In the synthesis process, when the image data extracted by the first extraction process and the image data extracted by the second extraction process are synthesized with an article separator as a boundary, an image related to the synthesized image data When the predetermined width in the main scanning direction exceeds the specified width, the image data extracted by the first extraction process and the image data extracted by the second extraction process are connected upstream of the predetermined direction. Align and compose with article separator,
Image processing device. The image processing apparatus according to any one of claims 3 to 6,
The controller is
In the end image search process, when the end image cannot be detected from the image data acquired by the image acquisition unit, the end image is searched from the image data acquired next by the image acquisition unit. To
Image processing device. The image processing apparatus according to any one of claims 1 to 7,
The controller is
Furthermore,
Processing for detecting the direction of characters included in the image related to the image data acquired by the image acquisition unit, and rotating the image with respect to the image data so that the direction of the character matches the direction of the source document Rotation process to execute
In the search process, the heading image and the end image are searched from the image data after the rotation process.
Image processing device. For each source document, an image acquisition unit that acquires image data, and information on column separators that classify images of source documents into predetermined column widths and the column widths between column separators are stored in association with each other. An image processing method executed by an image processing apparatus comprising:
A column separator search step for searching for the column separator based on the information of the column separator;
A search step for searching for a heading image and an end image from the image data acquired by the image acquisition unit;
An extraction step for extracting image data for each column region of the column width along the column separator from the heading image detected in the search step to the end image;
And a synthesis step of synthesizing the image data extracted for each column region by the extraction step. For each source document, an image acquisition unit that acquires image data, and information on column separators that classify images of source documents into predetermined column widths and the column widths between column separators are stored in association with each other. A program installed in a computer of an image processing apparatus comprising a storage unit,
In the computer,
A column separator search process for searching for the column separator based on the information of the column separator;
A search process for searching for a heading image and an end image from the image data acquired by the image acquisition unit;
From the heading image detected in the search process to the end image, an extraction process for extracting image data for each column area of the column width along the column separator;
And a synthesis process for synthesizing the image data extracted for each column area by the extraction process.

Images