JP5418064B2 - Image processing apparatus and program - Google Patents

Description

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

  As an apparatus for recognizing a table area in an image, for example, there is an apparatus disclosed in Patent Document 1. This apparatus identifies candidate areas of cell areas from an image, and selects combination candidates that can form a table for a plurality of candidate areas. Then, an evaluation value indicating the degree of possibility of configuring the table is calculated for each combination candidate, and the table having the maximum calculated degree is selected to configure the table.

JP 2007-213255 A

  An object of the present invention is to provide a technique for reducing a case where an area that is not a table area is determined as a table area.

An image processing apparatus according to claim 1 of the present invention includes an acquisition unit that acquires a binary image representing a binary image, and a graphic that identifies a graphic portion in the binarized image acquired by the acquisition unit. Specifying means; line segment specifying means for specifying a line segment in the part specified by the graphic specifying means; intersection determining means for determining an intersection relation of each line segment specified by the line segment specifying means; A cell structure in which rectangles are arranged in a plurality of rows or columns by the line segment specified by the segment specifying means and the intersection relationship determined by the intersection determining means, and an independent independent rectangle without forming the cell structure a structure specifying means for specifying, identified geometric portion in the previous SL diagram form specifying means based on the independent rectangle identified in the structural specification means is a means for determining whether a table region, A rectangle constituting the cell structure specified by the structure specifying means. Among the rectangles other than the rectangle with the independent rectangle specified by the structure specifying means inside, the ratio of the area of the rectangle in the area of the cell structure is calculated, and the figure is based on the calculated ratio. Table area determining means for determining whether or not the graphic part specified by the specifying means is a table area .

The image processing apparatus according to claim 2 of the present invention is the image processing apparatus according to claim 1 , wherein the structure specifying means further specifies the cell structure for the inside of the rectangle constituting the specified cell structure. Features.

An image processing apparatus according to a third aspect of the present invention is the image processing apparatus according to the first or second aspect , wherein the cell structure has a rectangular shape of 1 row n columns or n rows 1 column (n ≧ 2). It is characterized by a side-by-side structure.

An image processing apparatus according to a fourth aspect of the present invention is the image processing apparatus according to the second or third aspect , wherein the structure specifying unit is configured to determine a rectangular shape of the specified cell structure. The cell structure inside is specified.

According to a fifth aspect of the present invention, there is provided a program for identifying a computer with an acquisition unit that acquires a binarized image representing a binary image and a figure portion in the binarized image acquired by the acquisition unit. Figure specifying means, line segment specifying means for specifying line segments in the part specified by the figure specifying means, intersection determination means for determining an intersection relationship of each line segment specified by the line segment specifying means, A cell structure in which rectangles are arranged in a plurality of rows or columns by the line segment specified by the line segment specifying means and the intersection relationship determined by the intersection determining means, and independent independent without constituting the cell structure a structure specifying means for specifying a rectangle, the structure has been geometric portion identified in the previous SL diagram form specifying means based on the independent rectangle identified in particular means a unit that determines whether a table region The cell structure specified by the structure specifying means Among rectangles other than the rectangle in which the independent rectangle specified by the structure specifying means is inside, the ratio of the area of the rectangle in the area of the cell structure is calculated, and based on the calculated ratio And a program for causing the graphic portion specified by the graphic specifying means to function as a table area determining means for determining whether or not the graphic portion is a table area .

According to the inventions according to claims 1 and 5 of the present invention, it is possible to reduce an area that is not a table area from being determined as a table area, as compared with the case without the configuration of the present invention.
According to the invention according to claim 2 of the present invention, it is determined that a region that is not a table region is determined as a table region compared to a case where a cell structure is not further specified for the inside of a rectangle included in the specified cell structure. Can be reduced.
According to the invention of claim 3 of the present invention, an area that is not a table area is compared with a structure in which rectangles are arranged in 1 row and n columns or n rows and 1 column (n ≧ 2) as not specifying a cell structure. Can be reduced as a table area.
According to the fourth aspect of the present invention, a figure having a cell structure inside a corner of a rectangle included in the specified cell structure can be determined as a table region.

FIG. 3 is a block diagram showing a hardware configuration of the computer apparatus 1. The figure which showed the functional block implement | achieved in the computer apparatus. The flowchart which showed the flow of the process which CPU102 performs. The figure which showed an example of the binarized image. The figure which illustrated the format of line segment table TB10. The figure which showed the content of the list 300. FIG. The figure for demonstrating operation | movement of embodiment. The figure for demonstrating operation | movement of embodiment. The figure for demonstrating operation | movement of embodiment. The figure for demonstrating operation | movement of embodiment. The figure which showed an example of the binarized image. The figure for demonstrating operation | movement of embodiment. The figure which showed the content of the list 300. FIG. The figure which showed an example of the binarized image. The figure which showed the content of the list 300. FIG. The figure which showed an example of the binarized image. The figure which showed the content of the list 300. FIG. The figure which showed an example of the binarized image. The figure which showed the content of the list 300. FIG.

[Embodiment]
(Configuration of the embodiment)
FIG. 1 is a block diagram showing a hardware configuration of a computer apparatus 1 according to an embodiment of the present invention. In the present embodiment, the computer device 1 is a so-called PC (Personal Computer), and as shown in FIG. 1, each unit is connected to the bus 101, and exchanges information between each unit via the bus 101.

The display unit 107 includes a liquid crystal display as a display device for displaying an image, and displays a menu screen, various messages, images, and the like for operating the computer device 1 under the control of the CPU 102. The display device is not limited to a liquid crystal display, and may be another display device such as a display using CRT (Cathode Ray Tube) or EL (electroluminescence).
The operation unit 106 includes a keyboard and a mouse for operating the computer apparatus 1. When the user operates the keyboard and mouse, various instructions are input to the computer apparatus 1 and various settings are made.
The image acquisition unit 108 is an interface to which a scanner is connected, reads an image on a paper surface, communicates with a scanner that generates image data of the read image, and acquires image data from the scanner.

The storage unit 105 includes a hard disk device, and stores a program that causes the computer apparatus 1 to realize the function of the operating system, an image processing program that performs image processing on an image indicated by the image data, and image data acquired from a scanner. .
A ROM (Read Only Memory) 103 stores an IPL (Initial Program Loader), and a CPU (Central Processing Unit) 102 reads and executes the IPL stored in the ROM 103. When the IPL is executed, a program for realizing the function of the operating system is read from the storage unit 105 and executed. Further, when a program for realizing the function of the operating system is executed, the image processing program can be executed by the CPU 102. When the image processing program is executed by the CPU 102, the computer apparatus 1 functions as an image processing apparatus that processes the image indicated by the image data, and determines whether or not the figure in the image is a table. A function for specifying an area is realized.

FIG. 2 is a block diagram illustrating functions realized by the CPU 102 executing the image processing program.
The binarization unit 200 is a block that binarizes the image represented by the image data acquired by the image acquisition unit 108 into a black and white binary image. The binarized image (hereinafter referred to as a binarized image). It is an example of the acquisition means to obtain. In the present embodiment, the binarized image represents the pixel value by 1 bit, the black pixel is “1”, the white pixel is “0”, and the white pixel represents the background of the image. . In the binarized image, black pixels may be represented by “0” and white pixels by “1”. Further, the black pixel and the white pixel may be represented by numerical values other than “0” and “1”, such as “10” for the black pixel and “1” for the white pixel.
The graphic specifying unit 201 is a block that specifies a graphic part in the binarized image. As a technique for specifying a graphic portion from an image, for example, a known technique such as a technique described in Japanese Patent Laid-Open No. 3-102964 can be employed. Note that the technique for specifying a graphic portion from an image is not limited to the technique described in Japanese Patent Laid-Open No. 3-102964, and other techniques may be adopted.
The line segment specifying unit 202 is a block that specifies a line segment in the graphic part specified by the graphic specifying unit 201. As a technique for specifying a line segment, for example, as in the technique described in JP-A-2-210586, a binarized image is scanned, and black pixels are continuous beyond a predetermined threshold. Various known techniques such as a method of specifying a portion as a line segment can be employed. Note that the technique for identifying the line segment in the graphic portion is not limited to the technique described in Japanese Patent Laid-Open No. 2-258686, and other techniques may be employed.

The intersection determination unit 203 is a block for determining the intersection relationship between the line segments specified by the line segment specifying unit 202, and determines whether or not the line segments intersect from the coordinates of the start point and end point of the line segment. .
The structure specifying unit 204 is a block that specifies a rectangular grid portion in the graphic portion specified by the graphic specifying unit 201, and the line segment specified by the line segment specifying unit 202 and the line segments determined by the intersection determining unit 203 are Using the cross relationship, a rectangle in the graphic part is specified, and a structure in which the rectangles are arranged in a plurality of rows or a plurality of columns, an independent rectangle, etc. without configuring a cell structure are specified.
The list creation unit 205 is a block that generates a list representing the structure identified by the structure identification unit 204.
The table area determination unit 206 is a block that determines whether the graphic part specified by the graphic specification unit 201 is a table using the list generated by the list generation unit 205.

(Operation of the embodiment)
Hereinafter, an operation example of this embodiment will be described. In the following description, an operation when it is determined that the image indicated by the image data is not a table will be described first, and then an operation when it is determined that the image indicated by the image data is a table will be described.

(Operation when determining that an image is not a table)
First, when a user sets paper on which an image is drawn on a scanner connected to the image acquisition unit 108 and causes the image on the paper surface to be read, image data representing the image on the paper surface is generated in the scanner. Next, when the user performs an operation for instructing the acquisition of the image data generated by the scanner in the operation unit 106, the image acquisition unit 108 communicates with the scanner, and the image data generated by the scanner is acquired and stored. Stored in the unit 105.

  When the user performs an operation for instructing execution of the image processing program on the operation unit 106, the image processing program is executed by the CPU 102. Next, when the image data stored in the storage unit 105 is designated and the user performs an operation to instruct the designated image data to perform image processing on the operation unit 106, the CPU 102 follows the image processing program. The process shown in FIG. 3 is executed.

  Specifically, first, the CPU 102 performs a process of binarizing the image of the image data designated by the user (step SA1). Here, FIG. 4 is a diagram illustrating an example of the binarized image obtained by the binarization process, and the image G10 illustrated in FIG. 4 is a bar graph image. Since the CPU 102 binarizes the image data image, it functions as the binarization unit 200 here.

  When the binarization processing is completed, the CPU 102 analyzes the binarized image and specifies a graphic portion in the binarized image (step SA2). For example, in the case of the binarized image of FIG. 4, the part of the image G10 is specified as the part of the figure. When the CPU 102 specifies a graphic part in the binarized image, the CPU 102 generates a graphic identifier that uniquely identifies each graphic part, for example, generates a graphic identifier “1” for the part of the image G10. Since the CPU 102 specifies a graphic part in the binarized image, it functions as the graphic specifying unit 201 here.

  Next, the CPU 102 specifies line segments in the specified graphic part, and specifies the coordinates of each line segment (step SA3). For example, in the case of the binarized image in FIG. 4, line segments L101 to L110 drawn in the vertical direction and line segments L121 to L130 drawn in the horizontal direction are specified from the portion of the image G10. As the coordinates of the line segment, for example, the coordinates when the lower left corner of the image is the origin, the x coordinate is taken in the horizontal direction from the origin, and the y coordinate is taken in the vertical direction from the origin are specified. Since the CPU 102 specifies a line segment in the graphic part, it functions as the line segment specifying unit 202 here.

  When the CPU 102 specifies a line segment in each graphic part, the CPU 102 attaches a label number uniquely identifying each line segment to each line segment. A line identifier provided in a RAM (Random Access Memory) 104 is created by associating the graphic identifier attached to the graphic part with the label number of each line segment in the part specified by this graphic identifier and the coordinates of both ends of each line segment. Store in the table TB10 (step SA4). Here, FIG. 5 is an example of a format of the line segment table TB10. The line segment table TB10 has columns of figure identifiers, label numbers, coordinates of both ends of line segments, and label numbers of other line segments that intersect. In the present description, the number part in the code of each line segment is the same as the label number. For example, the label number 101 is the label number assigned to the line segment L101.

When the process of step SA4 is completed, the CPU 102 refers to each line segment to which a label number is attached and determines the crossing relationship between the line segments (step SA5). And the label number of the other line segment which cross | intersects one line segment is stored in line segment table TB10. That is, the CPU 102 functions as the intersection determination unit 203 that determines the intersection relationship between the line segments.
For example, in the case of the image of FIG. 4, the line segment L101 is associated with the label number “101” of the line segment L101 because the line segments L121, L122, L124, L128, and L130 intersect with each other. , L122, L124, L128, and L130 are stored as shown in FIG.

When the process of step SA5 is completed, the CPU 102 uses the information stored in the line segment table TB10 to specify a part having a rectangle in a plurality of rows or a plurality of columns in the specified graphic part (step SA6). A list 300 representing the relationship between the rectangles is created (step SA7). That is, here, the CPU 102 identifies a structure (cell structure) in which rectangles are arranged in a plurality of rows or columns, and a list creation unit 205 that creates a list representing the structure identified by the structure identification unit 204. Function as. In the present embodiment, when the CPU 102 specifies a portion having a rectangle in a plurality of rows or columns, the CPU 102 specifies a portion having a rectangle in n rows 1 column or 1 row n columns (n ≧ 2).
Specifically, first, the CPU 102 specifies a portion where rectangles are arranged in a plurality of rows or columns surrounded by a line segment that is the outer periphery of the graphic specified in step SA2, and labels each of the arranged rectangles. Perform processing to add. For example, in the case of the image in FIG. 4, the CPU 102 identifies line segments L101, L121, L110, and L130 as line segments that are the outer periphery of the figure from the coordinates of the line segments and the intersection relationship, and as shown in FIG. Then, the rectangle R10, the rectangle R20, the rectangle R30, and the rectangle R40 are specified as the portion having the rectangle. Then, the CPU 102 labels each rectangle with ID10, ID20, ID30, and ID40, and creates a list 300 in which each label is associated with the graphic identifier of the graphic as shown in FIG.
In the following description, when there are more rectangles in the specified rectangle, the rectangle included in the rectangle is referred to as a lower rectangle, and the rectangle including the rectangle inside is referred to as an upper rectangle. For example, since the portion of the rectangle R20 includes rectangles R21 to R23 described later, the rectangle R20 is an upper rectangle with respect to the rectangles R21 to R23, and the rectangles R21 to R23 are lower rectangles with respect to the rectangle R20. It becomes.

  For each specified rectangle, the CPU 102 further continues the process of specifying a portion having a rectangle in a plurality of rows or a plurality of columns in the rectangle. For example, for the rectangle R10, since there are no rectangles in a plurality of rows or columns in the rectangle, the CPU 102 ends the specification of the lower rectangle for the rectangle R10.

As for the rectangle R20, the CPU 102 identifies a plurality of rows of rectangles R21 to R23 as shown in FIG. 8, and labels each rectangle with ID21, ID22, and ID23, as shown in FIG. Each rectangular label is associated with ID 20 in list 300. Thereby, referring to the list 300 in FIG. 6, it can be seen that the rectangle specified by ID21, ID22, and ID23 is a lower rectangle of the rectangle specified by ID20.
Further, the rectangle R21 is not a plurality of rows or a plurality of columns, but has a rectangle inside. In this case, the CPU 102 specifies the rectangle inside the rectangle R21 as an independent rectangle (independent rectangle) that does not constitute a cell structure. Then, it is determined that the rectangle R21 including the independent rectangle is not a rectangle constituting a part of the table, and “?” Indicating that the rectangle R21 is not a rectangle constituting the table is ID21 as shown in FIG. Associate with.

As for the rectangle R30, as shown in FIG. 9, the CPU 102 identifies a plurality of rows of rectangles R31 to R37, labels each rectangle with ID31 to ID37, and sets each rectangle as shown in FIG. The rectangular label is associated with the ID 30 in the list 300.
As for the rectangle R34 and the rectangle R36, since there are a plurality of rows of rectangles in each rectangle, the CPU 102 labels each rectangle of the plurality of rows and, as shown in FIG. And ID 36 are associated with each rectangular label.
Note that the rectangle R37 has an independent rectangle inside rather than a cell structure. The CPU 102 assumes that the rectangle R37 having no cell structure is not a rectangle constituting a part of the table, and "?" Indicating that the rectangle R37 is not a rectangle constituting the table is displayed in the list 300 as shown in FIG. Corresponds to a certain ID37.

As for the rectangle R40, as shown in FIG. 10, the CPU 102 identifies a plurality of rows of rectangles R41 to R49, labels each rectangle with ID41 to ID49, and displays each rectangle as shown in FIG. The rectangular label is associated with ID 40 in the list 300.
As for the rectangle R48, since there are a plurality of rows of rectangles in the rectangle, the CPU 102 labels each rectangle of the plurality of rows, and as shown in FIG. Associate labels for each rectangle.

Next, the CPU 102 calculates the ratio of the area of each rectangle with a label to the area of the upper rectangle, and associates the calculated ratio with the label in the list 300 of FIG. 6 (step SA8).
For example, when the areas of the rectangles R10, R20, R30, and R40 are equal, the ratio of the area of each rectangle to the area of the image G10 assigned “1” as the graphic identifier is 25%. As shown in FIG. 6, “25%” is associated with ID10, ID20, ID30, and ID40 in the list 300. For example, the ratio of the area of the rectangle R21 to the area of the rectangle R20 is 60%, the ratio of the area of the rectangle R22 to the area of the rectangle R20 is 10%, and the ratio of the area of the rectangle R23 to the area of the rectangle R20 is 30%. In this case, the CPU 102 associates 60% with the ID 21 in the list 300, 10% with the ID 22, and 30% with the ID 23 as shown in FIG. If the ratio of the area is similarly obtained for the rectangles with other labels, the ratio of the area of each rectangle is associated with the label as shown in FIG. 6 in the case of the image shown in FIG. As described above, the CPU 102 also functions as the list creation unit 205 in order to create a list.

  Next, the CPU 102 identifies a label associated with “?” In the list 300, and replaces the identified ratio associated with the label with 0% (step SA9). For example, since “?” Is associated with ID21 which is the label of the rectangle R21, the CPU 102 replaces the ratio associated with ID21 with 0%. Further, since “?” Is associated with the ID 37 that is the label of the rectangle R 37, the CPU 102 replaces the ratio associated with the ID 37 with 0%.

When the CPU 102 finishes the process of step SA9, it determines whether or not the identified graphic part is a table based on the ratio associated with the labels in the list 300 (step SA10). That is, the CPU 102 functions as a table area determination unit 206 that determines whether or not the graphic part is a table.
Specifically, the CPU 102 first adds the ratio of the rectangular area to the lower rectangle, and if the addition result is less than a predetermined threshold, the CPU 102 sets the ratio stored in association with the upper rectangle to 0. Replace with%.
For example, for the rectangles R21 to R23 which are lower-layer rectangles with respect to the rectangle R20, the ratio associated with ID21 is “0%”, the ratio associated with ID22 is “10%”, and ID23. Since the associated ratio is “30%”, the addition result is 40%. Here, when the threshold value is 85%, for example, the addition result is 40% and is less than the threshold value. Therefore, for the upper layer rectangle R20, the ratio stored in association with ID20 is replaced with 0%. .

Next, for the rectangles of ID 341 and ID 342, the ratio associated with ID 341 is “80%”, the ratio associated with ID 342 is “20%”, and the addition result is equal to or greater than the threshold value. The ID 34 that is the label of the upper-layer rectangle R 34 is left as it is without replacing the associated ratio with 0%.
In addition, since “?” Is not associated with the rectangles whose labels are ID361 and ID362, the associated ratio is not replaced with 0% for ID36 which is the label of the rectangle R36 as the upper layer. Leave it as it is.
For the rectangles R31 to R37, the addition result is “65%” when the proportions associated with ID31 to ID37 are added. Here, since the addition result is 65% which is less than the threshold value, the ratio stored in association with ID30 which is the label of the upper layer rectangle R30 is replaced with 0%.
In addition, since “?” Is not associated with the rectangles of ID481, ID482, ID41 to ID49, the ratio associated with ID41 to ID49 and ID40 is not replaced with 0%.

  Next, the CPU 102 adds the ratios “25%”, “0%”, “0%”, and “25%” associated with the labels of the rectangles R10, R20, R30, and R40. Here, the addition result is 50%, and since the addition result is less than the threshold value, it is determined that the image G10 is not a table. When the CPU 102 obtains the determination result, the CPU 102 controls the display unit 107 to display the determination result on the display unit 107.

(Operation when determining that an image is a table)
Next, the operation when it is determined that the graphic portion of the binarized image is a table will be described. In this operation example, the operation will be described by taking as an example the case where the binarized image shown in FIG. 11 is obtained.

  First, in the present operation example, the same operation as the above-described operation example is performed from the time when image data is acquired until the intersection relationship between line segments in the binarized image is determined. When the processing from step SA1 to step SA5 is completed, the CPU 102 uses the information stored in the line segment table TB10 to specify a portion having a rectangle in a plurality of rows or a plurality of columns in the graphic portion (step SA6). A list 300 representing the relationship between the specified rectangles is generated (step SA7).

  For example, in the case of the image G11 shown in FIG. 11, as shown in FIG. 12, the rectangle R50, the rectangle R60, the rectangle R70, and the rectangle R80 are specified as a portion having a rectangle in a plurality of lines. Then, the CPU 102 labels each rectangle as ID50, ID60, ID70, and ID80. Also, labels are attached to the lower rectangles of the rectangle R50, the rectangle R60, the rectangle R70, and the rectangle R80, and a list 300 with each label is created as shown in FIG.

Next, the CPU 102 calculates the ratio of the area of each rectangle with a label to the area of the upper rectangle, and associates the calculated ratio with the label in the list 300 (step SA8). For example, when the areas of the rectangles R50, R60, R70, R80 are equal,
Since the ratio of the area of each rectangle occupying the area of the image G11 with “1” as the graphic identifier is 25%, the CPU 102 sets “25%” to ID 50 in the list as shown in FIG. , ID60, ID70, ID80. In addition, when the ratio of the area is similarly obtained for rectangles with other labels, in the case of the image shown in FIG. 11, the ratio of the area of each rectangle is associated with the label as shown in FIG.

  Next, the CPU 102 identifies a label associated with “?” In the list of FIG. 13, and performs processing to replace the ratio associated with the identified label with 0% (step SA9). In the case of the image shown in FIG. 11, since the list of FIG. 13 is created and there is no label associated with “?”, The proportion of the area associated with any label is 0%. Will not be replaced.

Next, the CPU 102 determines whether or not the identified graphic part is a table based on the ratio associated with the label in the list 300 (step SA10). When the CPU 102 adds the ratios “25%”, “25%”, “25%”, and “25%” associated with the labels of the rectangles R50, R60, R70, and R80, the addition result is 100%. Become. Here, since the addition result is equal to or greater than the threshold value, the CPU 102 determines that the image G11 is a table. When the CPU 102 obtains the determination result, the CPU 102 controls the display unit 107 to display the determination result on the display unit 107.
As described above, according to the present embodiment, an area that is not a table as in the graph is not recognized as a table, so that the recognition rate of the table can be increased.

[Modification]
As mentioned above, although embodiment of this invention was described, this invention is not limited to embodiment mentioned above, It can implement with another various form. For example, the present invention may be implemented by modifying the above-described embodiment as follows. In addition, you may combine each of embodiment mentioned above and the following modifications.

  When the outer frame is a non-rectangular figure as shown in FIG. 14, the CPU 102 that has executed the image processing program identifies the rectangle R101, the rectangle R102, the rectangle R103, and the rectangle R104 as a portion having a rectangle in a plurality of lines, and sets each rectangle. Even if ID101 to ID104 are labeled, the rectangle R105 and the rectangle R106 are specified as a portion having a rectangle in a plurality of columns, ID105 and ID106 are labeled, and the list 300 shown in FIG. 15 is created. Good. The list 300 shown in FIG. 15 is a diagram showing a list of states before associating the proportion of the rectangular area.

  In addition, in the case where the outer frame is not a rectangle as shown in FIG. 16, the CPU 102 that has executed the image processing program has a rectangle R201 and a rectangle R202, a rectangle R203 and a rectangle R204 as a portion having a rectangle in a plurality of lines. And the portions of the rectangle R205 and the rectangle R206 may be specified. The labels R201 and R202 are labeled ID201 and ID202, the rectangles R203 and R204 are labeled ID203 and ID204, the rectangles R205 and R206 are labeled ID205 and ID206, and the list 300 shown in FIG. 17 is created. Also good. Note that the list 300 shown in FIG. 17 is a diagram showing a list of states before associating the proportions of rectangular areas.

Further, the CPU 102 that has executed the image processing program specifies the rectangles R301 to R304 as a portion having a rectangle in a plurality of lines in the case of a figure having a rectangle at the corners inside the rectangle as shown in FIG. The rectangle R3041 and the rectangle R3042 may be specified inside the rectangle R304.
Then, the labels R301 to R304 may be labeled ID301 to ID304, the rectangles R3041 and R3042 may be labeled ID3041 and ID3042, and the list 300 illustrated in FIG. 19 may be created. The list 300 shown in FIG. 19 is a diagram showing a list of states before associating the proportions of rectangular areas.

In the present invention, the ratio of the area of the independent rectangle in the image may be calculated, and when the calculated ratio is equal to or greater than a predetermined threshold, it may be determined that the image is not a table.
In the present invention, the ratio of the area of the rectangle including the independent rectangle to the inside of the image is calculated, and when the calculated ratio is equal to or greater than a predetermined threshold, it is determined that the image is not a table. You may do it.

In the above-described embodiment, the threshold value may be set by the user by operating the operation unit 106.
Further, a plurality of threshold values may be set, and the user may select a threshold value from any one of a plurality of threshold values. For example, the two threshold values may be set so that one threshold value is, for example, 95% and the other threshold value is 50%.
Also, when setting the threshold, prepare multiple images of the table and graph, and use the learning results such as Widrow-Hoff learning rules and Fisher discriminant analysis based on the determination results for these images. You may make it set the threshold value at the time of determining whether it exists.

  In the above-described embodiment, the CPU 102 executes the image processing program so that the CPU 102 processes the image data and identifies the table area in the image data. However, the image processing described above is performed on the image of the image data. The configuration to be applied is not limited to this configuration. For example, a hardware such as a DSP (Digital Signal Processor) or a programmable logic device that performs the above-described processing is connected to the bus 101, and image data is supplied to the hardware so that the above-described processing is performed by the hardware. It may be.

  The above-described image processing program includes a magnetic recording medium (magnetic tape, magnetic disk (HDD (Hard Disk Drive), FD (Flexible Disk), etc.)), optical recording medium (optical disc (CD (Compact Disc)), DVD (Digital Versatile Disk). )), Etc.) may be provided in a state of being stored in a computer-readable recording medium such as a magneto-optical recording medium or a semiconductor memory and installed in the computer apparatus 1. Alternatively, it may be downloaded and installed via a communication line.

  In the embodiment described above, the PC executes the image processing program. However, a computer device other than the PC may execute the image processing program. For example, an image forming apparatus having a copying function or a scanner function may be used as an image. The image forming apparatus may execute the processing described above by executing a processing program.

DESCRIPTION OF SYMBOLS 1 ... Computer apparatus, 101 ... Bus, 102 ... CPU, 103 ... ROM, 104 ... RAM, 105 ... Memory | storage part, 106 ... Operation part, 107 ... Display 108: Image acquisition unit, 200: Binarization unit, 201 ... Graphic identification unit, 202 ... Line segment identification unit, 203 ... Intersection determination unit, 204 ... Structure identification , 205... List creation unit, 206... Table area determination unit

Claims (5)

Acquisition means for acquiring a binarized image representing a binary image;
Graphic specifying means for specifying a graphic part in the binarized image acquired by the acquiring means;
Line segment specifying means for specifying a line segment in a part specified by the graphic specifying means;
Intersection determination means for determining the intersection relationship of each line segment specified by the line segment specifying means;
A cell structure in which rectangles are arranged in a plurality of rows or columns by the line segment specified by the line segment specifying unit and the crossing relationship determined by the crossing determining unit, and independent without constituting the cell structure A structure specifying means for specifying an independent rectangle;
The identified geometric portion in the previous SL diagram form specifying means based on the independent rectangular identified a means for determining whether the tablespace by the structural identification means, identified in the structural specification means Of the rectangles constituting the cell structure, for the rectangles other than the rectangle with the independent rectangle specified by the structure specifying means inside, the ratio of the area of the rectangle to the area of the cell structure is calculated, and the calculated ratio An image processing apparatus comprising: a table area determining unit that determines whether or not the graphic part specified by the graphic specifying unit is a table area . The image processing apparatus according to claim 1, wherein the structure specifying unit further specifies the cell structure with respect to an inside of a rectangle constituting the specified cell structure. The cell structure, the image processing apparatus according to claim 1 or claim 2, wherein the rectangle on one line n columns or n rows and one column (n ≧ 2) is a structure aligned. The structure identification means, for rectangles constituting the identified cell structure, the image processing apparatus according to claim 2 or claim 3, characterized in that identifying the cell structure on the inside of the corners of the rectangle. Computer
Acquisition means for acquiring a binarized image representing a binary image;
Graphic specifying means for specifying a graphic part in the binarized image acquired by the acquiring means;
Line segment specifying means for specifying a line segment in a part specified by the graphic specifying means;
Intersection determination means for determining the intersection relationship of each line segment specified by the line segment specifying means;
A cell structure in which rectangles are arranged in a plurality of rows or columns by the line segment specified by the line segment specifying unit and the crossing relationship determined by the crossing determining unit, and independent without constituting the cell structure A structure specifying means for specifying an independent rectangle;
The identified geometric portion in the previous SL diagram form specifying means based on the independent rectangular identified a means for determining whether the tablespace by the structural identification means, identified in the structural specification means Of the rectangles constituting the cell structure, for the rectangles other than the rectangle with the independent rectangle specified by the structure specifying means inside, the ratio of the area of the rectangle to the area of the cell structure is calculated, and the calculated ratio A program for functioning as a table area determination means for determining whether or not the graphic part specified by the graphic specification means is a table area .

Images