JPH09106436A - Inclination detecting method for document picture - Google Patents

Abstract

PROBLEM TO BE SOLVED: To more accurately and efficiently detect an inclined angle by comparing only positions between patterns in the same shape. SOLUTION: A picture input means 101 such as a scanner inputs a picture. Next, a character pattern segmenting means 102 segments each character pattern from an inputted picture. Next, similar pattern detection part 103 compares these segmented patterns with each other to extract similar patterns to label them. When the labeling of the similar patterns is finished, the same line judging means 104 detects the similar pattern within the same line. Further, an angle between patterns detection means 105 detects an angle between the similar patterns within the detected same line. When the angle of each pattern pair is obtained like this, the whole angle deciding part 106 decides the angle of the whole picture.

Description

Detailed Description of the Invention

[0001]

The present invention relates to a method for detecting the inclination of a document image.

[0002]

2. Description of the Related Art As a conventional method for detecting the inclination of a document image, for example, a method of projecting features in various directions to detect a peak which is determined to be most along the line direction (Japanese Patent Laid-Open No. Hei 5). No. 174183)) and a method of obtaining the inclination by comparing the positions of circumscribed rectangles of the character patterns in the same line (see Japanese Patent Application Laid-Open No. 2-44486).

[0003]

The former method of projecting a feature in a plurality of directions as described above has a problem that a correct angle direction and a peak are formed depending on how the feature is taken.
In the latter method, it takes time to detect a line in a state where the inclination is unknown, and in addition, since different characters are originally compared, it is not possible to make a comparison at an accurate position, and the accuracy of inclination detection is high. There is a problem that it does not rise.

An object of the present invention is to first select the same pattern by the method for detecting the same pattern and compare the positions thereof, so that irrelevant character patterns are not compared with each other. An object of the present invention is to provide a document image inclination detection method for detecting inclination.

[0005]

In order to achieve the above object, according to the invention of claim 1, in a method for detecting a tilt of a document image, a pattern corresponding to a character is cut out from the document image, and the cut pattern is cut out. The feature is that the same label is given to similar patterns corresponding to the same character in the series, and the inclination of the image is detected based on the positional relationship between the patterns having the same label.

According to the second aspect of the invention, the inclination of the image is detected based on the positional relationship between the patterns of the same label existing in the same row.

According to the third aspect of the invention, based on the intersection of the line segment connecting the patterns of the same label and another pattern,
It is characterized by determining whether or not the patterns of the same label are present in the same row.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. FIG. 1 shows the configuration of an embodiment of the present invention. In the figure, 101 is an image input means such as a scanner, 102 is a character pattern cutout means, 103 is a similar pattern detection means for extracting a similar pattern from the cut out pattern sequence, and 104 is a similar pattern belonging to the same row. In-line determination means, 105 is means for detecting angles between detected similar patterns in the same row, 10
Reference numeral 6 is an angle determining unit that determines the angle of the entire image from the angles obtained by the similar patterns, 107 is a storage unit that stores various data, 108 is the overall control unit, and 109 is a data communication path.

FIG. 2 is a processing flowchart of the present invention. First, an image is input by the image input means 101 such as a scanner (step 201). Next, the character pattern cutout unit 102 cuts out each character pattern from the input image. This clipping is performed by, for example, integrating adjacent black pixel components. If the tilt angle is unknown, it is difficult to accurately extract characters, especially in Japanese. However, in the present invention, the character cutting is not the final purpose, and even if the character pattern is divided into biased or drastically, it does not pose a problem in the subsequent processing, so basically it is only necessary to extract the connected black pixel component.

Next, the similar pattern detecting means 103 compares the cut-out character patterns with each other, extracts similar patterns, and labels them (step 203). FIG.
Is a detailed flowchart of step 203. As the pattern matching method in step 303, for example, the test method described in paragraph 9 of US Pat. No. 5,303,313 is used. So this method
This is a method of superimposing patterns while correcting the position and determining whether or not they are similar depending on the position of the pixel in which a difference appears and the pattern of the pixel that appears. By this test method, when the pattern to be processed matches the existing template, the label attached to the matched template is added to the pattern to be processed (step 304). If they do not match, the pattern to be processed is registered as a new template and a new label is added (step 305). In step 302, in order to reduce the processing time, all the patterns may not be the processing targets, and only the patterns in the predetermined range in the image may be the processing targets.

After the labeling of the similar patterns is completed, the same line determination means 104 detects the similar patterns in the same line (step 204). FIG. 4 is a diagram for explaining detection of similar patterns in the same row, and FIG. 5 is a detailed flowchart of step 204.

First, a pair extracted as a similar pattern is selected (step 501). In FIG.
Let 402 be a similar pattern with the same label. When pattern matching is performed in the processing of step 203, the center position of the pattern is obtained in order to match the position of pattern superposition.
A line segment 404 connects the pattern center positions of the similar patterns 401 and 402. Further, 405 and 406 are line segments connecting the upper left apexes of the circumscribed rectangles of the pattern 401 and the pattern 402 and the lower left apexes of the circumscribed rectangles, respectively. Then, a rectangle intersecting the line segments 405 and 406 is detected (step 503). This rectangle is
In FIG. 4, they are 410, 411 and 412. Here, since the rectangle 411 is considerably smaller than the current processing target pattern 401, such a rectangle may be excluded from the processing target.

Next, of these intersecting rectangles, a rectangle that does not intersect the line segment 404 connecting the centers of similar patterns is detected (step 504). In the example of FIG. 4, rectangles 410 and 41
Since 1 and 412 intersect together, there is no corresponding one.
At this time, the pair of similar patterns 401 and 402 is determined to be in the same row, and is registered in the same row belonging pattern pair list (step 505).

Returning again to step 501, the next pair of similar patterns is selected. Here, it is assumed that the patterns 401 and 403 are similar patterns. A line segment 407 connects the pattern center positions of the similar patterns 401 and 403. Also, 408 and 409 are patterns 401 and 40.
It is a line segment connecting the upper left vertex and the lower left vertex of the circumscribed rectangle of No. 3. Line segments 408 and 40 as described above
A rectangle that intersects 9 is detected (step 503), and a rectangle that does not intersect the line segment 407 is detected (step 504). Since the rectangle 413 is detected as the corresponding rectangle, the pair of the pattern 401 and the pattern 403 is not set in the same row (step 506). By repeating the above process, it is possible to detect a pair of patterns existing in the same row.

Returning to FIGS. 1 and 2, the inter-pattern angle detecting means 105 detects the detected angle between similar patterns in the same row (step 205). In the example of FIG. 4, line segment 4
This is the angle of the line segment connecting the pattern centers such as 04. When the angle of each pattern pair is obtained, the overall angle determination means 106 determines the angle of the entire image (step 20).
6). This is the average value of the angles obtained in step 205,
The median is taken as the angle of the entire image.

Although the above-mentioned embodiment adopts a constitution having dedicated processing means, the present invention is not limited to this, and means 102 to 106 and 108 are
It is also possible to configure and process with one CPU.

[0017]

As described above, according to the first aspect of the invention, only the positions of the patterns having the same shape are compared, so that the tilt angle can be detected more accurately.

According to the second aspect of the present invention, since only the patterns in the same row are compared, the tilt angle can be detected more accurately.

According to the third aspect of the present invention, it is possible to determine whether or not the pattern exists in the same row by a simple process, so that the processing speed is improved.

[Brief description of the drawings]

FIG. 1 is a configuration of an embodiment of the present invention.

FIG. 2 is a processing flowchart of the present invention.

3 is a detailed flowchart of step 203 of FIG.

FIG. 4 is a diagram illustrating detection of similar patterns in the same row.

5 is a detailed flowchart of step 204 of FIG.

[Explanation of symbols]

 101 Image Input Means 102 Character Pattern Cutting Means 103 Similar Pattern Detecting Means 104 Same Line Judging Means 105 Inter-Pattern Angle Detecting Means 106 Overall Angle Determining Means 107 Data Storage Unit 108 Control Unit 109 Data Communication Channel

Claims (3)

[Claims] 1. A method of detecting a tilt of a document image, wherein a pattern corresponding to a character is cut out from the document image, the same label is given to similar patterns corresponding to the same character in the cut-out pattern series, and a pattern of the same label is given. A method for detecting the inclination of a document image, which comprises detecting the inclination of the image based on the positional relationship between them. 2. The skew detection method for a document image according to claim 1, wherein the skew of the image is detected based on a positional relationship between patterns of the same label existing in the same row. 3. A line segment connecting patterns having the same label,
3. The method for detecting the inclination of a document image according to claim 2, wherein it is determined whether or not the pattern with the same label exists in the same row based on the intersection with another pattern.