JP3517465B2 - How to order character areas - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic device such as an OCR (optical character recognition) device, a copying machine, and a facsimile, and in particular, a character ordering method for placing a reading order on a character area extracted from an input image. It is about.

[0002]

2. Description of the Related Art As a conventional ordering method, according to the position of each area in an original image, in the case of vertical writing, the order is from right to left, and similarly in the vertical direction, from top to bottom. There are things to go.

[0003]

However, in the above-mentioned prior art, when there are a plurality of articles in a manuscript such as a newspaper article, the texts of the plurality of articles are complicated and the reading order can be correctly set. However, there was a problem that corrections were necessary.

[0004]

In order to solve the above-mentioned problems, the ordering method of the present invention extracts a character area from an input image and sets the number of black pixels to the number of black pixels of the entire input image. On the other hand, according to the positional relationship between the character area and at least one of a separator image or a figure image having a ratio of a predetermined threshold value or more, the character area is divided into groups, and the order of the character areas is divided in each of the divided groups. decide. Further, in order to solve the above-mentioned problems, the ordering device of the present invention is an extraction unit that extracts a character region from an input image, and a black pixel number is predetermined with respect to the total number of black pixels of the input image. Grouping means for dividing the character area into groups according to a positional relationship between the character area and at least one of a separator image or a graphic image having a ratio of a threshold value or more, and an order of the character areas in each of the divided groups. And an order determining means for determining.

[0005]

With the above configuration, even if a plurality of articles are complicated in a manuscript, such as a newspaper article, the reading order can be correctly set, and the trouble of correction and the like can be reduced.

[0006]

[Embodiment] FIG. 20 is a block diagram showing the arrangement of an apparatus according to this embodiment. Reference numeral 1001 denotes a CPU that executes processing of the entire apparatus, and controls judgment and processing according to a control program stored in the ROM 1002. 1
A ROM 002 stores the control program of the flowchart described in the present embodiment, or data such as predetermined parameters used for processing. 1
A RAM 003 has a working memory area for storing data being processed by the CPU 1001. 10
A scanner 04 optically reads a document image.
The read image data can be stored in the RAM 1003. 1005 is a keyboard for inputting various codes,
Operator's instructions can be input. A pointing device 1006 can instruct a desired position on the display screen of the display 1007, and can also input a selection or cancellation instruction by clicking a button. 10
Reference numeral 07 is a display, which is a CRT or a liquid crystal display. Reference numeral 1008 denotes, for example, an LBP or inkjet printer, 1009 denotes, for example, an external storage device such as FD, 1
Reference numeral 010 is a data bus for exchanging data between these components.

[First Embodiment] FIG. 1 is a flowchart showing the ordering method of the present embodiment. In the figure, in step S101, an image is input from the scanner 1004 or the like.

Step S102 Next, the area of the image input in step S101 is divided, and areas such as characters, separators, and drawings are extracted from the image. The area dividing method in this step is, for example, Japanese Patent Application No.
There is a method disclosed in Japanese Patent Laid-Open No. 243252. All the area information divided here is stored in the RAM 1003.

Step S103 Next, area information relating to areas such as characters, separators and figures extracted from the image is read from the RAM 1003 and divided into hierarchical groups. The grouping will be described in detail below. FIG. 2 is a diagram showing a region extracted from an image, which is divided into six layers of the entire image (ROOT), article, column, same column group, paragraph (same line group), and line. Reference numeral 201 is layer 0, that is, the entire image. 202 is the article 1. Reference numeral 203 indicates that there are some more in the same hierarchy. The number of groups in the hierarchy is only one in hierarchy 0 and 1 in hierarchy 1 to 4.
There are more than one and there are more than one hierarchy 5. Reference numeral 204 is column 1 and belongs to article 1 (202). 205 is the same column group and belongs to column set 1 (204). 20
6 is a paragraph 1 and belongs to the same column group 1 (205). Paragraph 1 does not have a group hierarchy 5 of lines. 207 is a same-line group consisting of paragraph 2, that is, line 1 and line 2, and belongs to the same-column group 1 (205). Line 208 is in line 1 and belongs to paragraph 2 (207).

FIG. 3 is a detailed flowchart of step S103.

FIG. 4 shows a part of the character area, the separator area, the figure area, the articles created by the area division, the grouped article, the column group, the same column group, and the same line group. In FIG. 4, reference numeral 401 denotes a vertically written original image input from a scanner or the like. 413,443
Is an extracted character area, particularly a headline area and also an article, in-column group, or paragraph. Reference numerals 426 and 427 are extracted character areas, particularly heading areas and lines. 414, 41
Reference numerals 6, 430, and 444 are extracted character areas and in-column groups. 428 and 429 are extracted character areas and paragraphs. 415, 445, and 446 are extracted separator areas. 425 is an extracted drawing area. 41
1, 421 and 441 are the separator 446 and the drawing area 42.
It is an article area divided by 5. 412, 422,
442 is a column. 423 is a same-stage group.

FIG. 5 is a detailed flowchart of step S301. FIG. 6 is a detailed flowchart of step S304. FIG. 7 shows step S
3 is a detailed flowchart of 302. FIG. 8 is a diagram showing the regions extracted from the image of FIG. 4 divided into six layers of the entire image, article, column, same column group, paragraph (same line group), and line.
FIG. 9 is a detailed flowchart of step S104. FIG. 10 is a detailed flowchart of step S901. FIG. 11 is a detailed flowchart of step S902. FIG. 12 is a detailed flowchart of step S903. Figure 1
3 is a detailed flowchart of step S904. FIG. 14 is a detailed flowchart of step S905. FIG. 15 is a detailed flowchart of step S305. FIG. 16 shows the text area 426 extracted from the image of FIG. In FIG. 16, 1601 is a histogram of black pixels in the character area 426 in the Y direction (vertical direction). 1602 is the histogram 1
It is the width of the black pixel obtained from 601. FIG. 17 is a detailed flowchart of step S906.

Step S103 will be described with reference to the flowchart of FIG.

Step S301 A separator having a length of a certain ratio or more with respect to the entire image, the image is divided into one or more regions according to the drawing, and is made into an article. This will be described in detail with reference to FIG. 4 and the flowchart of FIG. First, in step S501, the area division S in FIG.
One separator or drawing extracted by 102 is taken out. Next, in step S502, the number of black pixels in the entire image is Lall, and the number of black pixels in the separator or figure is Lsep.
And

In step S504, it is determined whether or not Lsep ÷ Lall ≧ α (1) is satisfied. Here, α is a threshold value, which is 0.1 here. In step S505, the separator or diagram satisfying the expression (1) is stored in the array LONGSEP. If not satisfied, the process proceeds to step S506. In the example of FIG. 4, the separators 415 and 445 do not satisfy the formula (1). Since the separator 446 and the drawing area 425 satisfy Expression (1), they are stored in the array LONGSEP.

In step S506, it is determined whether all separators or figures have been processed. If all have been processed, the process proceeds to step S507. If there is any unprocessed separator or drawing, step S50.
Return to 1 to continue the process.

In step S507, the array LONGSE
The separators and figures stored in P are arranged in the order of longer width by comparing the ones having the longer widths or heights.

In step S508, the array LONGSEP is arranged.
Take out one of the separators or figures stored in. In step S509, the taken-out separator and the area are divided into two according to the drawing. In FIG. 4, first, the longest separator 446 divides the image 401 into left and right.
An area on the left side of the divided area is an article 441.

Next, in step S510, the array LONGS is arranged.
It is determined whether all separators and figures stored in the EP have been processed. If all have been processed, step S5
Proceed to 11. In the example of FIG. 4, since the drawing area 425 has not been processed, the process returns to step S508. The area on the right side formed by the division by the drawing area 425 is divided vertically. Then, the articles 411 and 421 of the layer 1 are created.

Since all the separators and figures stored in the array LONGSEP have been processed, step S511
Proceed to. In step S511, the frame and the table area are also separated as separate areas. For example, letters in newspaper columns, sentences in tables, data, etc., do not coexist with other text. Therefore, these also need to be treated as separate articles, so they are separated as separate areas. Then, the process proceeds to step S513.

In step S513, ID numbers are assigned to the divided areas as articles. In FIG. 4, the article 411 is referred to as the article 1, the article 421 is referred to as the article 2, and the article 441 is referred to as the article 3.

(Step S302) If the character areas extracted by the area division include areas consisting of only one line, they are combined into the same line group, and each region is made into a line. This will be described in detail with reference to FIG. 4 and the flowchart of FIG. First, step S7
At 01, retrieve one article. In the example of FIGS. 4 and 16, the article 421 having the article ID 2 is taken out and described.

Next, in step S702, one character area in the article is extracted. In step S703, it is determined whether the extracted area is a character area including only one line. If it is a character area including only one line, the process proceeds to step S704. If the character area is not composed of only one line, step S7.
Go to 05.

A method of checking how many lines the fetched character area consists of is shown in FIG. It is known from the result of area division in step S102 that the set direction of the character area 426 is vertical writing. When writing vertically, the character area 42
The vertical histogram of 6 black pixels is taken. When writing horizontally, the horizontal histogram is taken in the same way. In the example of FIG. 16, the histogram 1601 is the result of taking the histogram in the vertical direction. By looking at the histogram 1601 in the horizontal direction, it is possible to count how many consecutive black pixels exist, and the number of rows is obtained. Histogram 16
In the example of 01, the number of lines of the character area 426 is one because there is one continuous black pixel. Because it is one line,
It proceeds to step S704.

In step S704, there is a character area consisting of only one line that is adjacent in the same direction as the character area, and there is no separator or figure between them, and the size of the character is almost the same and the direction of the character is the same. Then, if the gap does not become large, use the same row group. In the example of the character area 426 of the article 421 in FIG. 4, the character area 426 is selected in step S703.
Is determined to consist of only one line, and it is written vertically, so a search is made for a character area consisting of only one line in the vertical direction. Character area 42
It can be seen that 7 is vertically aligned with the character area 426 and is composed of only one line. In addition, the character area 42
It is clear that there is no separator or figure between 6 and the character area 427. More specifically, since the two character areas are vertically aligned, the upper area that is aligned, in this case, the lower bottom of the character area 426, and the lower area that is aligned, in this case, the upper bottom of the character area 427. Check if part or all of the separator or figure is contained in the parallelogram consisting of. In the case of side-by-side alignment, check whether the separator or figure is included in the parallelogram that is also composed of the right side and the left side. Next, check the character size. In the example of FIG. 16, the width 1602 of the histogram 1601 is simply set to the character size. That is, the approximate character width is the character size. When a continuous area of black pixels is divided into several areas when a histogram is taken, that is, when there are multiple lines in a character area, the average of those widths is taken as the character size. The character sizes of the character region 426 and the character region 427 are compared, and the larger one is Cbig and the smaller one is Csmall.

If Csmall / Cbig ≧ α2 (2) is satisfied, the character sizes are the same. However, α2 is a threshold value and is 0.9 here. In the case of the character area 426 and the character area 427, since the expression (2) is satisfied, it is determined that the character sizes are the same. Next, check whether the assembly directions are the same. As described above, since the character area 426 is vertically written and the character area 427 is also vertically written, it is determined that the grouping directions are the same.

Next, it is examined whether the gap is large. Assuming that the gap between the character areas 426 and 427 is G and the smaller character size is Csmall, the gap between the character areas 426 and 427 is large if G ÷ Csmall ≤ α3 (3) is satisfied. It is determined not to. However, α3 is a threshold value and is set to 10 here. Therefore, the character area 426 and the character area 427
Are combined to form the same row group 424. Further, the character area 426 and the character area 427 are determined as lines.

In step S705, it is determined whether all character areas have been processed. If all have been processed, the process proceeds to step S706. If there is a character region that has not been processed, the process returns to step S702 to determine the same line group. The character areas 428, 429, and 430 are composed of a plurality of lines and thus do not form a line. In step S706, each row is given an ID number.

In step S707, it is determined whether all articles have been processed. If all have been processed, step S302 ends and the process proceeds to step S303. If there is an article that has not been processed, the process returns to step S701, and the same line group is determined.

Step S303 Here, the result of the area division in step S102 is directly assigned as an ID number as a paragraph of hierarchy 2. However, if the same line group generated in step S302 exists, it is similarly set as a paragraph of hierarchy 4 and an ID number is given. The area made into a line in step S302 is not a paragraph. In the example of FIG. 4, the same line group 424 generated from the character area 426 and the character area 427 is also treated as a paragraph.

Step S304 Here, a paragraph that constitutes one column of a column group is defined as a “same column group”. In the example of FIG. 4, paragraph 413 and paragraph 41
4, a paragraph 416, a paragraph (same line group) 424, a paragraph 430, a paragraph 443, and a paragraph 444 each form an in-column group of hierarchy 3 with only one area. Paragraph 42
8 and paragraph 429 belong to the same in-column group. Article 2 will be described according to the flowchart of FIG.
In step S601, one article is taken out. Here, an example of extracting article 2 is shown. In step S602,
Determines the grouping direction of articles. In the example of article 2 in FIG. 4, it is known from the result of the area division in step S102 that all of the character areas belonging to article 2 are vertically written, so that it is understood that the grouping direction of article 2 is vertically written. If vertical and horizontal character areas are mixed, the area of each is determined and the larger set direction is used. Alternatively, the direction with the larger number of each is set. Since it is written vertically, the process proceeds to step S603. In the case of horizontal writing, the process proceeds to step S609.

In step S603, the paragraphs in the article are arranged in descending order of height. In step S604, one paragraph that does not belong to any in-column group is taken out in the arranged order to form a basic in-column group. In step S605,
Other paragraphs that do not belong to any in-column group are taken out in the order in which they are arranged, and compared with the basic in-column group. If there are no separators or other paragraphs that belong to the same inline group, the areas overlap in the opposite direction to the writing direction of the article, and the gap is not too wide, the comparison paragraph is combined as the same inline group as the basic inline group. To do. Repeat other paragraphs that do not belong to any in-column group as a comparison.

In the example of article 2 in FIG. 4, the paragraph (same line group) 424 becomes the basic in-column group 424, but since the character size is different from that of the other paragraphs, the in-column group is formed independently. . Next, paragraph 428 becomes the basic paragraph. Between the comparison paragraph 429, the formula (2) is satisfied,
It is determined that the characters are the same size. In addition, there is neither a separator in the opposite direction (that is, a horizontal direction) to the set direction of articles nor a paragraph belonging to another in-column group. The determination method is the same as the method described in step S704. Next, the degree of overlap between the basic in-column group and the article in the comparison paragraph in the opposite direction (that is, in the horizontal direction) is determined. Letting Hsmall be the smaller height (width when writing horizontally) of the basic paragraph and the comparison paragraph, and let K be the overlapping length in the horizontal direction. If K ÷ Hsmall ≧ α4 ... (4) is satisfied, it overlaps. Determine that However, α4 is a threshold value and is 0.8 here. Basic in-column group 42
8 and comparative paragraph 429 satisfy equation (4). Next, check the gap. Since the gap between the basic in-column group 428 and the comparison paragraph 429 satisfies the formula (3), it is determined that the gap is not large. Therefore, the comparison paragraph 429 is referred to as the basic in-column group 42.
The same in-level group as 8 is combined to form one basic in-level group. Similarly, paragraph 430 is examined as a comparison paragraph. Since the region does not satisfy the formula (4) in the horizontal direction, it is not a same-stage group. Therefore, paragraphs 428 and 429
Only the same paragraph, the same column group 423
make.

In step S606, it is checked whether all paragraphs belong to any in-column group. In the example of article 2 in FIG. 4, since paragraph 430 still remains, step S6
Go to 04. Although paragraph 430 is a basic in-column group, there is no comparison paragraph in step S605, so paragraph 43
0 alone forms the same row group. Since all paragraphs belong to the same column group, the process proceeds to step S607.

In step S607, ID numbers are assigned to the same-stage groups. The paragraph 424 is the in-column group 1, the in-column group 423 is the in-column group 2, and the paragraph 430 is the in-column group 3.

In step S608, it is determined whether all articles have been processed. After processing all articles, step S30
4 is ended, and the process proceeds to step S305. If there are unprocessed articles, the process returns to step S601 to continue the processing.

[Step S305] Here, the same column group having the same column is collectively referred to as a "column". In the example of the article 411 and the article 421 in FIG. 4, the in-column groups 413, 424, and 443 each form a column group with only one in-column group. The same-column groups 414 and 416 form a column set 412. Similarly, in article 2, the same-column groups 423 and 430 have columns 40
Form 2.

Article 42 according to the flowchart of FIG.
1 will be described. In step S1501, one article is retrieved. Here, the article 421 is taken out. In step S1502, the set direction of articles is determined. This is the same as step S602. Here, the writing direction of articles is vertical writing. Since it is vertical writing, step S
Proceed to 1503. Similarly, in the case of horizontal writing, step S15
Go to 09. In step S1503, the in-column groups in the article are arranged in order from the right. In step S1504, one identical column group that does not belong to any column is taken out in the arranged order and is set as a basic column. In step S1505,
Other in-column groups that do not belong to any column are taken out in the order in which they are arranged, and compared to the same column group. If they overlap in the same direction and the gap is not too wide, combine the comparative same-column groups as the same column as the basic column. Other in-column groups that do not belong to any column are repeated as comparison in-column groups.

In the example of the article 421 in FIG. 4, first, the same column group 424 becomes the basic column set 424, but since the character size is different from other column sets, the basic column set is formed independently. Next, the same column group 423 becomes the basic column set 423. Expression (2) is satisfied between the comparison in-column group 430 and it is determined that the character sizes are equal.

Next, the degree of overlap of the articles in the same column group in the direction opposite to the grouping direction (that is, in the horizontal direction) is compared with the basic column. Assuming that the smaller height (width in horizontal writing) of the in-column group is Hsmall and the overlapping length in the horizontal direction is K, the equation (4) is satisfied.
It is determined that they overlap. It is also clear that they are next to each other. As for the checking method, as in step S704, it is checked whether there is another in-column group between them.

Next, the gap is examined. Since the gap between the basic column 423 and the comparative column group 430 satisfies the formula (3), it is determined that the gap is not large. Therefore, the comparative same-column group 430 is combined with the basic column as the same column as the basic column 423 to form the column 422. Since there is no other in-column group that should be the comparison in-column group, step S
The process ends 1505, and the process advances to step S1506.

In step S1506, it is checked whether all the same column groups belong to any column. If all the same column groups belong to the column set, the process advances to step S1507.

In step S1507, IDs are added to the columns.
Number them. The column set 424 is referred to as column set 1, and the column set 422 is referred to as column set 2. In step S1508, it is determined whether all articles have been processed. After processing all articles, step S30
5, the grouping is completed. If there are unprocessed articles, the process returns to step S1501 to continue the processing.

Thus, step S103 is completed. Figure 4
FIG. 8 is a diagram hierarchically showing the results of grouping in step S103 in the above example. In FIG. 8, each number corresponds to the number in FIG. Some layers have the same number in different layers, but this is because they form a single paragraph or in-column group by themselves. For example, column 1 of article 1
Is composed of a single in-column group 1, and further is composed of a single paragraph 1, that is, a character area 413, so that all have the same number 413.

Step S104 The ordering is performed for each group, and finally the ordering for the entire image is performed.

The ordering by group will be described in detail below. FIG. 9 is a detailed flowchart of step S104. FIG. 10 is a detailed flowchart of step S901. FIG. 11 shows step S
9 is a detailed flowchart of 902. 12
Is a detailed flowchart of step S903. FIG. 13 is a detailed flowchart of step S904. FIG. 14 is a detailed flowchart of step S905. FIG. 17 shows step S9.
It is a detailed flowchart about 06.

Step S104 will be described with reference to the flowchart of FIG.

(Step S901) All the rows in the same row group are ordered. This will be described in detail with reference to FIG. 4 and the flowchart of FIG. In step S1001, one same line group is taken out. In the example of FIG. 4, the same row group is only the same row group 424.

In step S1002, it is determined whether the set direction of the same line group is horizontal writing. Step S100 for horizontal writing
Go to 3. If it is vertical writing, the process proceeds to step S1004. The method of determining the set direction is the same as in step S602. Since the same line group 424 is composed of vertically written lines 426 and 427, it is determined to be vertically written, and the process proceeds to step S1004.

In step S1004, the rows included in the same row group are sequentially numbered from the top. Line 426
And row 427, row 426 is on top, so row 426, row 4
Number in order of 27.

In step S1005, it is determined whether all the same row groups have been processed. If all have been processed, step S901 ends and the process proceeds to step S902. If any remain, the process returns to step S1001 to continue the processing.
In the example of FIG. 4, since all processing has been performed, step S901 is ended and the process proceeds to step S902.

Step S902 The paragraphs and the same line groups in the same column group are all ordered. This will be described in detail with reference to FIG. 4 and the flowchart of FIG.

In step S1101, one in-column group is taken out. In the example of FIG. 4, since the in-column group 423 is the only in-column group formed of a plurality of paragraphs, the in-column group 423 will be mainly described.

In step S1102, it is determined whether the set direction of the in-column group is horizontal writing. Step S110 for horizontal writing
Go to 3. If it is vertical writing, the process proceeds to step S1104. The method of determining the set direction is the same as in step S602. The in-column group 423 includes vertical writing paragraphs 428 and 429.
Therefore, vertical writing is determined, and the process proceeds to step S1104.

In step S1104, the paragraphs and the same line groups included in the same column group are sequentially numbered from the right. In paragraphs 428 and 429, paragraph 428
Is right, so the numbers are given in the order of paragraph 428 and paragraph 429.

In step S1105, it is determined whether all in-column groups have been processed. If all have been processed, step S902 ends and the process proceeds to step S903. If any remain, the process returns to step S1101 to continue the processing.

(Step S903) The same column group in all articles is ordered. This will be described in detail with reference to FIG. 4 and the flowchart of FIG.

In step S1201, one column is extracted. In the example of FIG. 4, the column set 422 will be described.

In step S1202, it is determined whether the writing direction of the column is horizontal writing. If it is horizontal writing, the process proceeds to step S1203.
If it is vertical writing, the process proceeds to step S1204. The method of determining the set direction is the same as in step S602. The column 422 is
Vertical writing in-column group 423, in-column group 430
Therefore, vertical writing is determined, and the process proceeds to step S1204.

In step S1204, the same column groups included in the column set are numbered from the top. In the in-column group 423 and the in-column group 430, since the in-column group 423 is on the upper side, the in-column group 423 and the in-column group 430 are sequentially ordered.

In step S1205, it is determined whether all columns have been processed. If all have been processed, step S90
3 is ended and the process proceeds to step S904. If there are
It returns to step S1201 and continues processing.

Step S904 The columns in all articles are ordered. This will be described in detail with reference to FIG. 4 and the flowchart of FIG.

In step S1301, one article is taken out. In the example of FIG. 4, the article 421 will be described.

In step S1302, it is determined whether the writing direction of the article is horizontal writing. If it is horizontal writing, the process proceeds to step S1303.
If it is vertical writing, the process advances to step S1304. The method of determining the set direction is the same as in step S602. Article 421
Since it is composed of the vertical writing column 424 and the vertical column 422, it is determined to be vertical writing, and the process advances to step S1304.

In step S1304, the columns included in the article are sequentially numbered from the right. In the columns 424 and 422, since the column 424 is on the right side, the columns 424 and 422 are sequentially ordered.

In step S1305, it is determined whether all articles have been processed. If all have been processed, step S90
4 is ended and the process proceeds to step S905. If there are
The process returns to step S1301 to continue the processing.

[Step S905] Articles are ordered respectively. This will be described in detail with reference to FIG. 4 and the flowchart of FIG.

In step S1401, it is determined whether the set direction of the entire image is horizontal writing. If it is horizontal writing, the process proceeds to step S1402. If it is vertical writing, the process proceeds to step S1403. The method of determining the set direction is the same as in step S602. Image 401
Is composed of a vertically written article 411, an article 421, and an article 441. Therefore, it is determined to be vertically written, and the process proceeds to step S1403.

In step S1403, the articles included in the entire image are numbered in order from the right, and in step S9.
05 is ended and the process proceeds to step S906. Article 41
The order is 1, the article 421, and the article 441.

Step S906 All character areas of the entire image are ordered as a whole. This will be described in detail with reference to FIG. 4 and the flowchart of FIG.

In step S1701, 1 is assigned to the variable ORDER (initialization). This variable value is the order for the entire image of the character area. In step S1702, one article is taken out from the smallest number. Here, the number means each I assigned in steps S901 to S905.
D number. In step S1703, step S17
The columns in the article extracted in 02 are extracted in ascending order of number. In step S1704, step S1703
The same column group in the column set extracted in step 1 is taken out in ascending order of number. In step S1705, step S1
The paragraphs in the in-column group extracted in step 704 are extracted in order from the smallest number. In step S1706, it is determined whether the paragraphs extracted in step S1705 are in the same line group. If it is the same row group, the process advances to step S1707. If not in the same row group, step S1708
Proceed to. In step S1707, one line in the paragraph extracted in step S1705 is extracted in ascending order of number. In step S1709, the order of the variable ORDER is assigned to the row, and 1 is added to ORDER. In step S1710, it is determined whether all rows have been processed. If all rows have been processed, the process advances to step S1711. If any remain, the process returns to step S1707 to continue the processing. Also, in step S1708, the paragraph is assigned the order of the variable ORDER, and 1 is added to ORDER. In step S1711,
Determine if all paragraphs have been processed. If all paragraphs have been processed, the process advances to step S1712. If there are
The process returns to step S1705 to continue the process. Step S
At 1712, it is determined whether all in-column groups have been processed. If all in-tier groups have been processed, the process advances to step S1713. If so, step S1704
Return to and continue processing. In step S1713, it is determined whether all columns have been processed. If all columns have been processed, the process advances to step S1714. If any remain, the process returns to step S1703 to continue the processing. Step S171
At 4, it is determined whether all articles have been processed. If all articles have been processed, step S906 ends. If any remain, the process returns to step S1702 to continue the processing. Thus, step S104 is completed.

[Second Embodiment] As shown in the flowchart of FIG. 5, in the processing relating to the grouping of articles in step S301 of FIG. 3, in the first embodiment, step S
Although the threshold value α of 504 is set to a constant value, it may be changed according to the number of black pixels of the image as shown in step S1901 of FIG. For example, when the number of black pixels in the image is large, the threshold value is set small, but as the number of black pixels in the image decreases, the threshold value is increased. As a result, since the length of the separator or the like is actually small, but the image is also small, it is possible to reduce mistakes such as separating a portion that does not need to be separated, and improve the ordering accuracy. For example, in the case of an A4 400 DPI image, the number of black pixels in the entire image is 100,000 or more α = 0.1 50,000-100,000 α = 0.05 10,000-50,000 α = 0 Less than 0.03 10,000 α = 0.01 Similarly, the threshold values used in the equations (2) to (4) may be changed depending on the number of black pixels of the image, the layer, and the like. The flowchart showing the ordering method of this embodiment is the same as that of the first embodiment shown in FIG. With the above description, the same operation and effect as those of the above-described first embodiment can be obtained.

[0073]

As described above, according to this embodiment, even if there are a plurality of articles in a manuscript such as a newspaper article, the reading order can be correctly set, and the time and effort for corrections can be reduced. There is an effect that can be done.

[Brief description of drawings]

FIG. 1 is a flowchart showing an ordering method according to this embodiment.

FIG. 2 is a diagram showing areas extracted from an image, which are divided into six layers.

FIG. 3 is a detailed flowchart of step S103.

FIG. 4 is an exemplary diagram of each area extracted by area division.

FIG. 5 is a detailed flowchart of step S301.

FIG. 6 is a detailed flowchart of step S304.

FIG. 7 is a detailed flowchart of step S302.

FIG. 8 is a diagram showing areas extracted from the image of FIG. 4 divided into six layers.

FIG. 9 is a detailed flowchart of step S104.

FIG. 10 is a detailed flowchart of step S901.

FIG. 11 is a detailed flowchart of step S902.

FIG. 12 is a detailed flowchart of step S903.

FIG. 13 is a detailed flowchart of step S904.

FIG. 14 is a detailed flowchart of step S905.

FIG. 15 is a detailed flowchart of step S305.

16 is a diagram showing a character area 426 extracted from the image of FIG.

FIG. 17 is a detailed flowchart of step S906.

FIG. 18 is a diagram showing the areas extracted from the image of FIG. 4 divided into three layers.

FIG. 19 is a detailed flowchart of step S301.

FIG. 20 is a configuration block diagram of a device according to the present embodiment.

Front page continued (58) Fields surveyed (Int.Cl. ⁷ , DB name) G06K 9/00-9/82 H04N 1/40 G06T 1/00 G06T 7/00-7/60 G06F 17/20-17 / 26

Claims (4)

(57) [Claims] 1. A character area is extracted from an input image, and at least one of a separator image or a graphic image having a ratio of the number of black pixels to the total number of black pixels of the input image is a predetermined threshold value or more. An ordering method, characterized in that the character areas are divided into groups according to a positional relationship with the character areas, and the order of the character areas is determined in each of the divided groups. 2. The ordering method according to claim 1, wherein the predetermined threshold value is determined according to the number of black pixels of the entire input image. 3. The ordering method according to claim 1, wherein when the character areas are divided into groups, the character areas are divided into groups having a plurality of layers. 4. Extraction means for extracting a character area from an input image, and a separator image or a graphic image having a ratio of the number of black pixels to the total number of black pixels of the input image being a predetermined threshold value or more. According to a positional relationship between at least one of the character areas and the character area, grouping means for dividing the character area into groups, and order determining means for determining the order of the character areas in each of the divided groups, Ordering device.