JPH06208625A - Image processing method and apparatus - Google Patents

Abstract

(57) [Summary] (Correction) [Purpose] To improve the accuracy when segmenting the input image data for each attribute of a document, and once the segmentation is performed and the attribute is identified, the attribute To facilitate the process of correcting [Structure] A rectangle is recognized by inputting image information and detecting consecutive pixels for the black pixel, and using the width, height, area, and pixel density of the region indicated by the recognized rectangle. By temporarily identifying the attributes such as characters, figures, and tables, and calculating and comparing partial histograms of the recognized regions, the attributes of the regions are determined.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and apparatus capable of dividing input image information into areas such as character areas, figures and tables.

[0002]

2. Description of the Related Art In a conventional image processing apparatus, as a region dividing means, a table and other regions are classified based on the ratio of the area of the circumscribed rectangle and the number of pixels to determine whether or not the region is a table region, and based on the result. The method was to divide the area.

Further, when the attributes (characters, figures, tables, etc.) of the divided areas are erroneously identified in the conventional image processing apparatus, the erroneously identified area data is eliminated, and the operator manually operates again. It was to specify the correct area frame.

[0004]

However, in the above-mentioned conventional technique, when a separator made of complicated polygonal lines or the like (deformed separator) is treated as one region, the number of pixels for the area of the circumscribed rectangle becomes large. There is a problem that the region is erroneously recognized as a region and the region division accuracy is lowered.

Further, in the prior art in which the divided areas are corrected manually by the operator, the burden on the operator increases when there are a plurality of erroneous divided areas, and a large amount of time is spent as a ratio to the processing time of the entire image processing. There was a problem that became.

[0006]

In order to solve the above-mentioned problems, the present invention recognizes a rectangle by inputting image information and detecting continuous pixels with respect to black pixels of the image information, The width of the area indicated by the recognized rectangle,
Identify the attributes of the rectangle using height, area and pixel density,
Calculating a partial histogram of the recognized region,
There is provided an image processing method and apparatus characterized by determining the attribute of the area by comparing.

In order to solve the above-mentioned problems, the present invention inputs image information, recognizes a rectangle by detecting continuous pixels with respect to black pixels of the image information, and recognizes a rectangle by the recognized rectangle. The attribute of the rectangle is identified by using the width, height, area, and pixel density of the indicated area, the ruled line of the recognized area is detected, and the attribute of the area is determined by the number of the ruled lines. An image processing method and device are provided.

In order to solve the above-mentioned problems, the present invention inputs image information, recognizes a rectangle by detecting consecutive pixels with respect to black pixels of the image information, and recognizes the rectangle. The attribute of the rectangle is identified using the width, height, area, and pixel density of the area indicated by, and the attribute of the area is determined from the ratio of the area of the area identified as the table area and the area in the table frame. There is provided an image processing method and apparatus characterized by determining

In order to solve the above problems, preferably the attribute is the text.

In order to solve the above problems, the attribute is preferably a graphic.

In order to solve the above-mentioned problems, preferably the attribute is a table.

In order to solve the above-mentioned problems, the present invention inputs image information, divides the input image information into regions for each attribute, stores the region information and the attribute information obtained by the region division, and stores the information. There is provided an image processing method and apparatus characterized by instructing correction of the attribute information stored and updating the stored attribute information in accordance with the instruction.

[0013]

【Example】

[Embodiment 1] FIG. 1 is a block diagram showing an image processing apparatus of this embodiment.

In the figure, reference numeral 101 is an image data input unit.

Reference numeral 102 denotes an arithmetic processing unit (CPU) for controlling the apparatus and performing arithmetic operations for processing of each unit according to a control program stored in the storage unit 103.

Reference numeral 103 denotes a storage unit for storing a control program for processing as shown in a flowchart described later and various data, including a ROM and a RAM.

An output unit 104 outputs the arithmetic processing result, the image processing result, and the image data.

Reference numeral 105 denotes an area dividing unit showing the entire area dividing processing.

Reference numeral 106 denotes a labeling section for labeling the pixels of the input image and simultaneously creating initial rectangular data.

Reference numeral 107 is an attribute detection unit for detecting attributes of separators, tables, figures, and the like.

Reference numeral 108 denotes a table ruled line detecting section for detecting straight lines forming a table.

Reference numeral 109 is a table area determining unit for determining whether or not the table area is a table area based on the state of the table ruled line.

FIG. 2 is a flow chart showing the image processing in the image processing apparatus of this embodiment. The control program for executing this processing is stored in the storage unit 103.

Step S201 First, the original image is input from the image input unit 101.

Step S202 Next, in the labeling unit 106, labels are added to the black pixels of the thinned image one line at a time, the same label is given to pixels that are continuous vertically, horizontally, and diagonally, and at the same time, a rectangle is shaped. Go.

Taking FIG. 3 as an example, the first detected pixel A is labeled 1, and the coordinates (Xa, Xa,
Ya) is the start and end points of the rectangle, the number of pixels is 1, and the same label 1 as the pixel is added to the rectangle label for distinguishing the rectangle, and the above data is stored in the storage unit as rectangle data (FIG. 4).

Next, a label 2 is attached to a pixel B having no continuous pixels in the left direction (there is no continuous pixel from the top because it is the first row), and the coordinates (Xb, Yb) of this pixel A are rectangular. The number of pixels is 1, the same label 2 as the pixel is added to the rectangular label for distinguishing the rectangle, and the above data is also stored in the storage unit as rectangular data (FIG. 4).

When the labeling of the first line is completed as described above, the process moves to the second line.

The first pixel C in the second row is the pixel A of label 1.
Since it is continuous from the top, pixel label 1 is added, and the number of pixels is added to the rectangular data of rectangular label 1 by 1 to give a total of 2 pixels, and the rectangular label remains 1, and the rectangular coordinates are only the end point. Update from (Xa, Ya) to (Xa, Yc) (the coordinates of the start point do not change).

Since the next pixel D is continuous with the pixel C from the left, the label 1 is added, and the number of pixels is added to the rectangular data of the rectangular label 1 by 1 to make a total of 3 pixels, and the rectangular label remains 1. Until now, only the end point of the rectangular coordinates (Xa, Yc)
To (Xd, Yc) (the Y coordinate of the end point does not change).

At this time, the pixel D is diagonally continuous with the pixel B and is continuous from the pixel C. Therefore, the label of the pixel B is changed from the label 2 to the label 1, and the rectangular data of the rectangular label 1 is obtained. On the other hand, the total number of pixels of the rectangular label 2 is 4
The rectangle label remains the same as 1 as a pixel, and the rectangle coordinates are set to the end point (X) so that the rectangle coordinates include all the pixels A, B, C, and D.
(d, Yc) to (Xb, Yd). The rectangular data of the rectangular label 2 is invalidated by setting the rectangular label to 0.

When the second line is completed as described above, the process moves to the third line.

Since the first pixel E in the third row is diagonally continuous with the pixel C, the pixel label 1 is added, and the number of pixels is increased by 1 to the rectangular data of the rectangular label 1, resulting in a total of 5 pixels, which is a rectangle. The label remains unchanged at 1, and the rectangular coordinates are updated from the starting point (Xa, Ya) to (Xe, Ya) and the ending point (Xb, Yd) to (Xb, Ye).

In the same manner, labeling and rectangular tracing are performed for all thinned pixels.

Step S203 Next, in the attribute detection unit 107, after the labeling and the tracing of the rectangle, the rectangle corresponding to the text, the rectangle corresponding to the figure or the photograph, the table, the rectangle corresponding to the separator, etc. , The height H, the area S, and the number of pixels with respect to the area, that is, the pixel density D (these are easily obtained by calculation using rectangular data).

In the separator, the width W is not more than the threshold Tw1 and the height H is not less than twice the threshold Tw of the width W (S50 in FIG. 5).
3) or the width W is larger than the threshold Tw1 and the height H
Is 3 times the threshold Tw of the width W or more (S504 in FIG. 5), the rectangular label is unified as -3 as the vertically long separator,
Pixel labels forming the rectangle are retained as they are, apart from the rectangle label (S514 in FIG. 5).

Also, in the case where the width and the height are interchanged, the rectangular label is changed to -3 as the horizontally long separator in the same manner as above (S505 and S506 in FIG. 5).

Next, as shown in FIG. 6, the pixel density D is equal to the threshold value T.
When d1 or less (S507 in FIG. 5), the rectangular label is changed to -3 by considering it as a deformed separator area such as a key shape (S514 in FIG. 5).

When the area S is larger than the threshold value Ts1 (S508 in FIG. 5), if the pixel density D is less than the threshold value Td2 (S509 in FIG. 5), it is regarded as a table and the rectangular label is changed to -4 (FIG. 5). If the pixel density D is greater than or equal to the threshold value Td2, the rectangular label is changed to -5 as if it is a figure or a photograph (S516 in FIG. 5).

Further, when the area S is not more than the threshold Ts1 and not less than the threshold Ts2 (S510 in FIG. 5), the pixel density D is not less than the threshold Td3 (S511 in FIG. 5), and the width W and the height H are both. When the threshold value Tw4 or more and the pixel density D is Td5 or more (S513 in FIG. 5), the rectangle is regarded as a figure or a photograph and the rectangle label is changed to -5.

Further, when the area S is less than or equal to the threshold Ts1 and greater than or equal to the threshold Ts2 (S510 in FIG. 5), the rectangle having the pixel density D less than the threshold Td4 (S512 in FIG. 5) is regarded as a table and the rectangle label is set to -4. It is changed (S515 in FIG. 5).

As described above, a rectangle corresponding to a figure or photograph, a table, etc., a rectangle corresponding to a separator, etc. are detected,
With the remaining rectangle as the body, the rectangle label is made to be the same as the pixel label (S517 of FIG. 5).

Step S204 Next, the table area determining unit 108 determines the width for calculating the histogram from the width W and the height H of the rectangle determined to be the table area (S901 in FIG. 9).

Only the pixels (701 in FIG. 7 and 801 in FIG. 8) corresponding to the table frame excluding the text and the like (702 in FIG. 7 and 802 in FIG. 8) inside the rectangle determined to be the table area (this is the rectangle (It is easily obtained by comparing and selecting the pixel label of the data and the pixel label of each pixel inside the rectangle), and a partial histogram at the left and right ends in the vertical direction and the horizontal direction is calculated (S902 in FIG. 9), and the rectangle Histogram threshold value Tkh indicating whether the table frame line is parallel to the Y axis from the height of H, and Histogram threshold value Tk indicating whether the table frame line is parallel to the X axis from the width of the rectangle
w is determined (S903 and 904 in FIG. 9).

Next, when there are portions corresponding to the table frame lines at the left and right ends of both the horizontal and vertical partial histograms (that is, when there are table frame lines at the left and right ends and the top and bottom ends of the rectangle). The rectangle corresponding to () is determined to be a table area (S905, S906, S908 in FIG. 9).

If at least one of the left and right edges of the horizontal and vertical partial histograms is missing the portion corresponding to the table frame line (that is, either the left edge or the right edge or the upper edge or the lower edge). If it is missing, the relevant rectangle is determined as the modified separator area (S in FIG. 9).
905, S906, S907).

Step S205 Finally, the rectangular data of various areas obtained as described above is output from the output unit 104 together with the image data.

It should be noted that, at the time of inputting the image in step S201, it is judged whether the image is multi-valued and converted into a binary image.
Region division processing is possible even if the input image is a multi-valued image such as a color image.

When the image is input in step S201,
If the number of pixels of the input image is large enough to require a great amount of processing time, it is possible to speed up the processing by performing image thinning processing for converting an image of vertical m dots and horizontal n dots into one pixel.

Further, in the detection of the separator in step S203, the rectangular labels are distinguished by the vertical / horizontal difference of the separator, the difference in the threshold value when obtaining figures, tables, etc., so that a more detailed area attribute classification can be performed. It is also possible to do it.

When determining the table area in step S204, the differences Sx and Sy between the left end and the right end and the upper end and the lower end histogram are respectively taken as shown in FIGS. 10 and 11, and there is either one of the differences. If the length is equal to or larger than the threshold value Ts, it is determined that the modified separator region is not uniform in ruled line length, and if it is less than the threshold value Ts, it is determined that the deformation region is the table region.

Further, at the final output of step S206, it is possible to refer to the rectangular data of various areas and output only the image data of the necessary area, and further reduce the storage area and process. The time can be further shortened.

[Second Embodiment] In this embodiment, the first embodiment
As a method of confirming the table area, the histograms of the top, bottom, left, and right edges of the area whose attribute was determined to be a table were taken, whereas the histogram of the entire area determined to be a table was taken, the ruled line was detected, and this number of table areas An example of determining that is will be described.

A block diagram showing the configuration of the image processing apparatus of this embodiment is as shown in FIG.

The processing of this embodiment is shown in the flowchart of FIG. 12, but the same steps as those in the flowchart of FIG. 2 in the first embodiment are designated by the same step numbers, and the description thereof will be omitted here.

Step S1204 Next, in the table ruled line detecting unit 108, the pixels (701 in FIG. 7) corresponding to the table frame excluding the text and the like (702 in FIG. 7 and 802 in FIG. 8) inside the rectangle determined to be the table area. , 80 in FIG.
For only 1) (this can be easily obtained by comparing and selecting the pixel label of the rectangular data and the pixel label of each pixel inside the rectangle), the histogram is calculated in the vertical direction and the horizontal direction (S901 in FIG. 9), Histogram threshold value Tkh whether the ruled line is parallel to the Y axis from the height of the rectangle, and histogram threshold Tk whether the ruled line is parallel to the X axis from the width of the rectangle
w is determined (S902, 903 in FIG. 9).

Next, from the shape of the histogram in the vertical direction, a place which is equal to or more than the threshold value Tkh is regarded as a ruled line of the table area (9 in FIG.
04) Similarly, from the shape of the histogram in the horizontal direction, a place equal to or larger than the threshold value Tkw is regarded as a ruled line in the table area (9 in FIG.
05).

Step S1205 Next, in the table area determination unit 109, the number Nx of vertical ruled lines obtained as a ruled line from a histogram in the vertical direction and a part having a threshold value Tkh or more and the part having a threshold value Tkw or more from the histogram in the horizontal direction are determined. The number Ny of horizontal ruled lines obtained as ruled lines is stored, and if there are three or more ruled lines, the table area is determined (S906, S908, S90 in FIG. 9).
9) If the number of ruled lines is less than 3, the modified separator area is defined (S906, S908, S907 in FIG. 9).

Step S1206 Finally, the rectangular data of various areas obtained as described above is output from the output unit 104 together with the image data.

It should be noted that when the image is input in step S201, it is judged whether it is multi-valued and is converted into a binary image, so that even if the input image is a multi-valued image such as color, the area division processing is possible. is there.

When the image is input in step S201,
If the number of pixels of the input image is large enough to require a great amount of processing time, it is possible to speed up the processing by performing image thinning processing for converting an image of vertical m dots and horizontal n dots into one pixel.

Further, in the detection of the separator in step S203, the rectangular labels are distinguished by the vertical / horizontal difference of the separator, the difference of the thresholds when obtaining figures, tables, etc., so that more detailed area attribute classification can be performed. It is also possible to do it.

When it is known that the inclination of the image after input is small at the time of detecting the table ruled line in step S1204, the table ruled line can be detected by tracing the line instead of calculating the histogram.

Further, when the table area is determined in step S1205, instead of fixing the threshold value for determining whether it is a table area as shown in FIG. 16, the number of ruled lines defined corresponding to the size of the rectangular area is changed. It can be substituted by comparing with certain thresholds Tn1 and Tn2, if there are ruled lines equal to or larger than the threshold value, it is decided as the table area, and if there are less ruled lines than the threshold value, the modified separator area is used.

Further, as shown in FIG. 17, a threshold value determined corresponding to the size of the area of the rectangle which is tentatively determined to be the above table area is determined from the width and height of the rectangle, and the threshold value of the number of vertical ruled lines is set to TnH.
Alternatively, the threshold value of the number of horizontal ruled lines may be set to TnW for substitution.

Further, at the final output of step S1206, it is also possible to refer to the rectangular data of various areas and output only the image data of the necessary area, and further reduce the storage area and process. The time can be further shortened.

[Third Embodiment] In this embodiment, the first embodiment
In the second embodiment, the method of determining the table area is to take the histogram of the area whose attribute is determined to be a table, whereas the table area is determined by the ratio of the area in the table frame to the area of the area determined to be a table. An example of determining that is will be described.

A block diagram showing the configuration of the image processing apparatus of this embodiment is as shown in FIG.

The processing of this embodiment is as shown in the flowchart of FIG. 2, but step S204 is different from that of the first embodiment and will be described here.

Step S204 Next, the table area deciding unit 108 decides the table area. FIG. 22 is a flow chart showing in detail the table area confirmation processing in step S204. Description will be given below with reference to the flowchart of FIG.

First, in step S221, the outline of the outer frame of the table is traced to store the point coordinates forming the outline.

Here, FIG. 23 shows steps S221 to S221.
It is a figure explaining the process of 223.

In FIG. 23, reference numeral 2301 denotes a rectangular area determined to be a table (very small), white circles are white pixels, black circles are black pixels, and gray circles are white pixels, which are pixels inside the table,
Is shown.

The black pixels and the like in the table are omitted.

First, the inside of the table area is searched line by line from above from the left to search for black pixels forming the table frame. When a black pixel is found, its coordinates are stored in the storage unit 103. In the example of the area 2301, it is checked whether or not there is a black pixel for each pixel in the direction of coordinates (0, 1) to (10, 1). (2, 1)
Since there is a black pixel in, the coordinates are stored in the storage unit 103.

Next, the contour line is traced clockwise or counterclockwise from the black pixel, and the coordinates of the black pixels (hereinafter referred to as contour points) forming the contour are stored in the storage unit 103. Area 230
In the example of No. 1, (2, 1)-(10, 1), (1
0, 1)-(10, 11), (10, 11)-(1, 1
1), (0, 10)-(0, 4), (1, 4)-(1,
Up to 2) are stored in the storage unit 103 as contour points.

Next, in step S222, the number of pixels belonging to the inside of the table is counted and added for each horizontal line. Here, for each line, the two rightmost and leftmost contour points stored in the storage unit 103 in step S221 are extracted, and the distance between the two extracted points is added as the number of pixels belonging to the inside of the table. . When expressed by an equation, (x-coordinate of right end) − (x-coordinate of left end) + 1 = (number of pixels in table).

By doing so, the area inside the table can be easily and quickly obtained even if the contour line is complicated or the table frame is missing. This is sufficient for the confirmation processing performed in the table area confirmation unit 108, but of course the area inside the table may be accurately calculated.

In the example of the area 2301 in FIG. 23, first, consider the line having the y coordinate of 1, which is the uppermost horizontal line. y
In the line of = 1, the leftmost contour point is (2, 1) and the rightmost contour point is (10, 1). Therefore, the number of pixels in the table is 10-2 + 1 = 9.

In step S223, it is determined whether the pixels of all the lines in the table area have been counted. If all lines have been completed, the process proceeds to step S224. If there are still lines remaining, the process returns to step S222 to continue the processing.

In the example of the area 2301, since counting of all the lines has not been completed yet, the process returns to step S222 and the line of y = 2 is considered. On the y = 2 line, the leftmost contour point is (1, 2) and the rightmost contour point is (1
0, 2).

Therefore, the number of pixels in the table is 10-1 + 1 = 10. This is added to the number of pixels 9 calculated on the line of y = 1, and the total is 19.

Similarly, in the example of the area 2301, the process is performed up to y = 11. Therefore, the number of pixels in the table is 9 + 10 + 10 + 11 + 11 + 11 + 11 + 11 + 11
+ 11 + 10 = 116.

In step S224, the ratio of the area inside the frame to the area of the rectangular area determined to be a table is calculated.

Next, in step S225, it is determined whether the ratio of the area inside the frame is larger than the threshold value. That is, (area in frame) / (area of rectangular area determined to be a table)>
If (threshold value γ) is satisfied, the corresponding rectangle is determined as the table area in step S226, and the process proceeds to step S205. If not satisfied, the corresponding rectangle is determined as a modified separator in step S227, the rectangular label of the rectangle stored in the storage unit 103 is changed to -3, and the process proceeds to step S205. However, here, the threshold value γ is set to γ = 0.9.

In the example of the area 2301, (area inside the frame) = 9 + 10 + 10 + 11 + 11 + 11 + 1
1 + 11 + 11 + 11 + 10 = 116, (area of rectangular area determined to be table) = 11 × 11 = 12
1, therefore, 116 ÷ 121 = 0.959> 0.9, and (area in frame) ÷ (area of rectangular area determined to be table)>
Since (threshold value γ) is satisfied, the area 2301 is determined as the table area, and the process proceeds to step S205.

Similarly, in the example shown in FIG. 7, the inside of the table surrounded by the frame 702 is represented as a black portion 801 in FIG. 8, and the ratio of the area within the frame becomes a value larger than the threshold value γ. Therefore, the table is confirmed and the process proceeds to step S205.

In FIG. 18, 701 indicates a rectangular area determined to be a table, 702 indicates a frame, and 703 indicates a line of a horizontally written sentence.

Similarly, in the example shown in FIG. 20, the inside of the table surrounded by the frame 902 is represented as a black portion 1001 in FIG. 21, and the ratio of the area within the table frame becomes a value smaller than the threshold value γ. Therefore, it is determined as a modified separator.

In FIG. 20, reference numeral 901 denotes a rectangular area determined to be a table, 902 a frame, and 903 a line of a vertically written sentence.

When the table area is determined in step S204,
Although the number of pixels belonging to the inside of the table is counted and added for each horizontal line in step S222, it can be replaced by counting and adding the number of pixels belonging to the inside of the table for each vertical line.

Further, when the table area is determined in step S204, the number of pixels belonging to the inside of the table is counted and added for each horizontal line in step S222. At that time, black pixels forming the text in the frame are added. The table area finalizing process can be performed with either counting or not counting.

Further, when the table area is determined in step S204, the number of pixels belonging to the inside of the table is counted and added for each horizontal line in step S222 to obtain the area inside the frame. Even if the area inside the frame is calculated by the above, the table area confirmation processing is possible.

At the final output in step S205, it is also possible to refer to the rectangular data of various areas and output only the image data of the necessary area, and further reduce the storage area and process. The time can be further shortened.

[Embodiment 4] In this embodiment, for example, a process of correcting a divided area obtained by the method described in Embodiments 1 to 3 and an attribute of the area will be described.

FIG. 24 is a block diagram showing the arrangement of the image processing apparatus of this embodiment.

In FIG. 24, the configuration of 101 to 105 is the same as that shown in FIG. 1, and therefore its explanation is omitted here.

Reference numeral 111 denotes a display unit for displaying the result of the division by the region division unit 105 in a frame by superimposing it on the original image data and displaying a list of attributes.

Reference numeral 112 denotes an area division correction unit for correcting erroneous division of the area divided by the area division unit 105.

Reference numeral 113 is an area attribute correction unit for correcting the area attribute estimated by the area dividing unit 105.

FIG. 25 is a flow chart showing the area division correction processing in the image processing apparatus of this embodiment. The control program for executing this processing is stored in the storage unit 103.

FIG. 28 is a flow chart showing the area attribute correction processing in the image processing apparatus of this embodiment, and the control program for executing this processing is stored in the storage unit 103. When an area that should be one is accidentally divided into multiple areas.

In the area division correction unit 112, a plurality of areas erroneously divided by the area division unit 105 are surrounded by a frame (301 in FIG. 26) with a pointing device or the like (S2501 in FIG. 25). At that time, the display of the corresponding area frame (area A, area B, area C in FIG. 26) is changed by changing from a line to a dotted line (S2503 in FIG. 25).

However, if there is an area that extends beyond the instruction frame, the instruction frame is invalidated and erased, and the process returns to the beginning (S2502 in FIG. 25).

If all the areas fit within the designated frame, the area data (coordinates, attributes, etc., FIG. 27) of all the areas enclosed by the frame are stored in the storage unit 103 (S25 of FIG. 25).
04), at the same time, a new area attribute setting list (30 in FIG. 26).
3) is displayed (S2505 in FIG. 25).

Next, by selecting the attribute of the new area from the list (S2506 of FIG. 25), the new attribute is highlighted (S2507 of FIG. 25). If the new attribute is determined (S2508 in FIG. 25), any one of the plurality of areas is randomly selected and the coordinates, attributes, etc. of the area data are changed (S2509 in FIG. 25), and the other remaining areas are changed. A new area frame (302 in FIG. 26) is displayed as invalid data (S2510 in FIG. 25) (S2511 in FIG. 25).

By the above operation, a plurality of areas are combined into one.

When the attribute of the area is wrong: In the area attribute correcting section 113, when it is desired to change only the attribute of the area data estimated by the area dividing section 105,
Only the target area (602 in FIG. 29) is designated by enclosing it with a frame (603 in FIG. 29) with a pointing device or the like (S2801 in FIG. 28), and a list of attributes (60 in FIG. 29).
1) is displayed (S2802 of FIG. 28). At that time, the current attribute is highlighted (S2803 in FIG. 28) and the display of the corresponding area frame is changed from a line to a dotted line (602 in FIG. 29) to clearly indicate (S2804 in FIG. 28).

Next, by selecting the attribute to be changed from the list (S2805 in FIG. 28), the reverse display of the original attribute is erased and the new attribute is displayed in reverse (S in FIG. 28).
2806), at the same time, the attribute of the area data is changed (FIG. 2).
8 S2807).

In the area division correction unit 112, an alternative means can be used by moving the cursor up / down / left / right with a key on the keyboard instead of instructing with a pointing device or the like.

Further, instead of using the instruction frame to indicate that a plurality of regions have been selected by displaying the frame of the region by dotted lines, an alternative means can be displayed by changing the display color of the region frame.

In addition to the example of FIG. 27, the area data structure may be a structure including data such as the number of pixels in the area and the set direction of the text.

It is also possible to process only the merging of the areas first, and to independently perform the attributes of the areas later.

In the area attribute correction unit 113, an alternative means can be used by changing the display color instead of the reverse display in representing the current attribute of the attribute list.

Further, instead of changing the display of the corresponding area frame from the line to the dotted line, the color of the frame can be changed as an alternative means.

Still, instead of displaying a list of attributes, only the current attributes are displayed, and the display of the attribute display portion is changed by providing an instruction such as a pointing device or a display change switch to display the desired attribute. By repeating up to, it can be used as an alternative method for changing attributes.

[0117]

As described above, the input image is converted into rectangular data and classified into areas such as a character area, a figure or a photograph, a table, and a separator, and the vertical direction of the area temporarily determined to be the table area and By dividing partial histograms on the left and right edges in the horizontal direction and comparing them to determine whether they are table areas, it is possible to divide complicated area configurations without changing existing electronic parts, circuits, etc. It is possible to obtain an improvement in accuracy with respect to the division of the area configuration. Processing time is short because a histogram is partially taken. And so on.

As described above, the input image is converted into rectangular data and classified into character areas, figures or photographs, tables, separators, etc., and the vertical and horizontal directions of the areas temporarily determined to be table areas are further classified. The existing electronic parts,
There is an effect that the precision can be improved and the like especially for the division of the complicated area structure without changing the circuit etc., especially for the division of the area structure including the table.

As described above, the input image is converted into rectangular data and classified into character areas, figures or photographs, tables, separators, and the like, and the area within the frame with respect to the area of the area provisionally determined to be the table area. By determining the ratio and determining that it is a table when the ratio is greater than or equal to the threshold value, it is possible to divide a complicated area structure, especially an area structure including a table, without changing existing electronic parts, circuits, etc. However, the accuracy is improved. Since the area inside the frame is taken, the table can be accurately determined even if the table frame is slightly missing. Since it takes the area inside the frame, it is strong against inclination and the table can be accurately determined.

There are advantages such as the following.

As described above, by combining erroneously divided areas and by correcting the erroneously estimated attributes, the operability is improved without changing the existing area division processing.
This has the effect of reducing the processing time.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image processing apparatus according to a first embodiment.

FIG. 2 is a flowchart showing image processing in the image processing apparatus of the first embodiment.

FIG. 3 is a diagram showing an example of labeling processing according to the first embodiment.

FIG. 4 is a diagram showing a relationship between a rectangular data structure and rectangular labels and attributes according to the first embodiment.

FIG. 5 is a flowchart showing the processing of the separator etc. detection unit of the first embodiment.

FIG. 6 is a diagram showing division of attribute (density) D and area s of a rectangle (region) according to the first embodiment.

FIG. 7 is a diagram showing a histogram when the table area of Example 1 is determined.

FIG. 8 is a diagram showing a histogram when the modified separator region of Example 1 is determined.

FIG. 9 is a flowchart illustrating a processing example of a table area determining unit according to the first embodiment.

FIG. 10 is a diagram showing a histogram when the modified separator area of Example 1 is determined.

FIG. 11 is a flowchart illustrating a processing example of a table area determining unit according to the first embodiment.

FIG. 12 is a flowchart showing image processing in the image processing apparatus according to the second embodiment.

FIG. 13 is a diagram showing a histogram when the table area of Example 2 is determined.

FIG. 14 is a diagram showing a histogram when the modified separator region of Example 2 is determined.

FIG. 15 is a flowchart illustrating a processing example of a table ruled line detection unit and a table area determination unit according to the second embodiment.

FIG. 16 is a flowchart illustrating a processing example of a table area determining unit according to the second embodiment.

FIG. 17 is a flowchart illustrating a processing example of a table area determining unit according to the second embodiment.

FIG. 18 is a diagram showing a rectangle indicating a table and a table area according to the third embodiment.

FIG. 19 is a diagram showing a rectangle indicating the inside of the table and the area of the table in the third embodiment.

FIG. 20 is a view showing a modified separator of Example 3 and a rectangle showing a region of the modified separator.

FIG. 21 is a view showing a rectangle showing the inside of the modified separator and the area of the modified separator of Example 3;

FIG. 22 is a flowchart illustrating a processing example of a table area determining unit according to the third embodiment.

FIG. 23 is a diagram showing a table area and each pixel in Example 3;

FIG. 24 is a block diagram showing a schematic configuration of an image processing apparatus of Example 4.

FIG. 25 is a flowchart showing a region division correction process according to the fourth embodiment.

FIG. 26 is a diagram showing an example of area division correction processing according to the fourth embodiment.

FIG. 27 is a diagram showing a region data structure according to the fourth embodiment.

FIG. 28 is a flowchart showing region attribute correction processing according to the fourth embodiment.

FIG. 29 is a diagram showing an example of a region attribute correction processing display unit according to the fourth embodiment.

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁵ Identification number Internal reference number FI Technical indication location G06K 9/38 C (72) Inventor Akihiro Matsumoto 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Within the corporation

Claims (14)

[Claims] 1. Inputting image information, recognizing a rectangle by detecting consecutive pixels with respect to black pixels of the image information, width, height of an area indicated by the recognized rectangle, Identifying the attributes of the rectangle using area and pixel density, calculating a partial histogram of the recognized region,
An image processing method characterized by determining the attribute of the area by comparing. 2. The image information is input, a rectangle is recognized by detecting consecutive pixels with respect to the black pixels of the image information, and the width and height of the area indicated by the recognized rectangle are detected. An image processing method characterized in that the attribute of the rectangle is identified by using the area and the pixel density, the ruled line of the recognized region is detected, and the attribute of the region is determined by the number of the ruled line. 3. The image information is input, a rectangle is recognized by detecting consecutive pixels with respect to the black pixels of the image information, and the width and height of the area indicated by the recognized rectangle are detected. Image processing characterized in that the attribute of the rectangle is identified using the area and the pixel density, and the attribute of the area is determined from the ratio of the area of the area identified as the table area and the area in the table frame. Method. 4. The image processing method according to claim 1, wherein the attribute is a text. 5. The image processing method according to claim 1, 2, or 3, wherein the attribute is a graphic. 6. The image processing method according to claim 1, wherein the attribute is a table. 7. An input means for inputting image information, a rectangle recognition means for recognizing a rectangle by detecting consecutive pixels for black pixels of the image information, and a rectangle recognized by the recognized rectangle. Attribute identifying means for identifying the attribute of the rectangle using the width, height, area, and pixel density of the area, and calculating a partial histogram of the recognized area,
An image processing apparatus, comprising: a control unit that controls to determine the attribute of the area by making a comparison. 8. An input unit for inputting image information, a rectangle recognition unit for recognizing a rectangle by detecting consecutive pixels for black pixels of the image information, and a rectangle recognized by the recognized rectangle. Attribute identifying means for identifying the attribute of the rectangle using the width, height, area and pixel density of the area, ruled line detecting means for detecting the ruled line of the recognized area, and attribute of the area depending on the number of the ruled lines And a control means for controlling so as to fix the image processing apparatus. 9. An input unit for inputting image information, a rectangle recognition unit for recognizing a rectangle by detecting consecutive pixels with respect to black pixels of the image information, and a rectangle recognized by the recognized rectangle. Attribute identifying means for identifying the attribute of the rectangle using the width, height, area, and pixel density of the area, and the area of the area identified as the table area and the ratio of the area in the table frame to the area An image processing apparatus, comprising: a control unit that controls to determine an attribute. 10. The image processing apparatus according to claim 7, 8 or 9, wherein the attribute is a text. 11. The image processing apparatus according to claim 7, wherein the attribute is a graphic. 12. The image processing apparatus according to claim 7, wherein the attribute is a table. 13. Inputting image information, dividing the input image information into regions by attribute, storing the region information and the attribute information obtained by the region division, and instructing correction of the stored attribute information, An image processing method, characterized in that the stored attribute information is updated in accordance with the instruction. 14. Input means for inputting image information, area dividing means for dividing the input image information into areas for each attribute, storage means for storing the area information and the attribute information obtained by the area division, and the storage. An image processing apparatus comprising: a correction instructing means for instructing correction of the generated attribute information; and an update control means for updating the stored attribute information in accordance with the instruction.