JP2001109888A - Ruled-line extraction system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a ruled-line extraction system which can perform a ruled- line extracting process corresponding to the quality of an image. SOLUTION: An image input means 101 obtains an input image and image generating means 102 with different resolutions generates a low-resolution image and a high-resolution image. A ruled-line candidate area extracting means 103 extracts a ruled-line candidate area by using the generated low-resolution image. A quality evaluating means 104 for an image searches the pixels in the extracted ruled-line candidate area to evaluate the quality of the image and a means 105 which selects a processing method or threshold according to the quality selects a processing method or threshold matching the image quality according to the result of the evaluation by the quality evaluating means 104. A means 106 for selecting image resolution by partial processes selects an image to be processed according to the image quality. Through the means mentioned above, the proper processing method, threshold, and image to be processed for a ruled-line extracting means 17 are selected to extract a ruled line.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a ruled line extraction method, and more particularly to a ruled line extraction method for extracting ruled lines corresponding to image quality in a character recognition device such as an OCR.

In recent years, with the spread of input devices such as scanners, the demand for a character recognition device (or software) OCR has been increasing. In order to realize a high recognition rate of each character, a clipping process, which is a stage before recognition, is important in terms of its accuracy. Heretofore, when processing a form or the like by OCR, it has been necessary to specify information such as the position of ruled lines and the thickness of lines in advance and store information about a range in which characters are written. Further, it is necessary to input information every time the form format is changed, which is a burden on the user. The present invention is a technique for solving such a problem, and accurately forms a ruled line of a table-shaped frame for a form image whose position and format are unknown, and whether the quality of the image quality is good or bad. This is a method that can be extracted.

[0003] It should be noted that the present technology is applicable not only to character recognition devices but also to line recognition in drawing recognition, ruled lines in images and objects,
The present invention can be applied to a technique for extracting a straight line in a broad sense, such as separation of a contact portion between a figure and a character.

[0004]

2. Description of the Related Art In the conventional OCR, it is necessary to store information such as ruled line position information and line thickness in advance as form data. However, there is a problem that a form in an unknown format cannot be processed.

Therefore, the present applicant has previously filed Japanese Patent Application No. 9-50 / 1990.
At 527, a processing method that does not require frame information in advance was proposed. In the method described here, as shown in FIG. 44, a straight line is extracted by a horizontal straight line extracting unit 4403 and a vertical straight line extracting unit 4410 from the partial pattern extracted by the connected pattern extracting unit 4401, and the horizontal straight line extracting unit 4
The horizontal straight line extracted in 403 is grouped two by two from the top by the horizontal frame determining unit 4417 to form a horizontal frame, and the vertical straight line extracting unit 4417.
A vertical frame is determined by the vertical frame determining unit 4418 from the vertical straight lines extracted in step 10, and a rectangular range surrounded by two horizontal frames and two vertical frames is extracted as a frame by the rectangular expression unit 4422. Then, it is assumed that the ruled line candidates forming the frame are the original ruled lines, and the ruled line candidates not forming the frame are patterns other than the ruled lines, and are not ruled lines.

[0006]

However, when the above-described processing is actually performed, it is necessary to target various low-quality images including blurring and crushing. Image data often includes both high-quality images and poor-quality images. In the present circumstances,
In each case of good or bad image quality, it is possible to set appropriate processing or thresholds.

However, when the image is set as a high-quality image, the accuracy of character recognition may be reduced if the image includes a part of low quality. In addition, even when the image is set as a low quality image, if the image includes a part of high quality, the accuracy of character recognition may be reduced.

For example, when a high-quality image is set, processing using a low-resolution image is performed. However, in a processing method using only a low-resolution image, a relatively crushed image is processed. The effect of blurring is reduced, and the accuracy is high for blurred images.
For other images, the accuracy is reduced due to the influence of the collapse.

When an image is set as a low-quality image, processing using a high-resolution image is performed. However, in a processing method using only high-resolution images, a high-quality image The accuracy can be improved because the image can be examined in detail. However, for a blurred image, the influence of the blurring becomes large, so that the accuracy becomes low.

Therefore, even if the image is of poor quality, good quality, or a mixture of these images, the ruled lines and the four sides are surrounded with ruled line candidates with high precision. There is a need for a technology that can automatically extract a cell region that is a rectangular region that has been formed.

An object of the present invention is to provide a ruled line extraction method for performing a ruled line extraction process corresponding to the quality of an image.

[0012]

In order to solve the above-mentioned problems, according to the first aspect of the present invention, means for generating images of different resolutions from an input image, means for evaluating the quality of the input image, Means for selecting an image having a resolution corresponding to the quality from among the images having different resolutions; and means for extracting ruled lines or cells from the selected image. This makes it possible to select a suitable image and extract ruled lines based on the evaluation result of the quality of the input image, thereby improving the accuracy of ruled line extraction.

According to a second aspect of the present invention, there is provided a ruled line extraction method for extracting a ruled line forming a form in image data, comprising: means for generating images having different resolutions from an input image; Means for evaluating the quality of the input image using the images generated by the means for generating images of different resolutions, and extracting ruled lines or cells based on the quality evaluated by the means for evaluating the quality of the input image Means for selecting a processing method or a threshold value used for the processing, means for selecting an image resolution suitable for processing for extracting ruled lines or cells from respective portions of the input image, and selecting the selected image A processing method for extracting the ruled line or the cell from an image of resolution or a processing method selected by a unit for selecting a threshold used in the processing Other comprises based on a threshold, and means for extracting a ruled line or cell constituting a form, a. This makes it possible to select a suitable image and extract ruled lines based on the evaluation result of the quality of the input image, thereby improving the accuracy of ruled line extraction.

According to the third aspect of the present invention, the means for generating images having different resolutions from the input image includes a low-resolution image having a large amount of data to be thinned out from the input image and a high-resolution image having a small amount of data to be thinned out from the input image. create. This makes it possible to use an image with an appropriate resolution in subsequent processing.

According to a fourth aspect of the present invention, the means for evaluating the quality of the input image expresses the quality by a numerical value. This makes it possible to quantify the image quality.

According to a fifth aspect of the present invention, the image processing apparatus further comprises a ruled line candidate area extracting means for extracting a ruled line candidate area before performing the processing by the image quality evaluating means. This makes it possible to calculate the image quality of the required part.

According to the invention of claim 6, the ruled line candidate area extracting means performs a ruled line candidate extracting process using a low resolution image. This makes it possible to more accurately extract ruled line candidate positions even when the image is blurred.

Further, according to the present invention, before performing the processing by the ruled line candidate area extracting means, a process of extracting a dotted line using a low resolution image and removing a pattern constituting the dotted line is performed. . By using a low-resolution image, even when the image is blurred, ruled line candidate positions can be more accurately extracted, and a portion that is known as a dotted line can be reliably removed.

According to the eighth aspect of the present invention, the ruled line candidate area extracted using the low resolution image is subjected to the dotted line extraction processing and the ruled line extraction processing using the high resolution image, whereby the attribute of the ruled line is obtained. Judge. This makes it possible to more accurately extract even a dotted line with a short interval between points that are likely to be crushed in a low-resolution image, and processing with higher accuracy even if the dotted line is a line candidate by mistake. Can be performed.

According to the ninth aspect of the present invention, the means for evaluating the quality of the image uses an image having a different resolution (higher resolution image) from the image used in the ruled line candidate area extracting means. Evaluate the quality. As a result, it is possible to eliminate the influence of the crushing that occurs on the low-resolution image.

According to the tenth aspect of the present invention, in the means for improving the quality of the image, when the inside of the ruled line candidate area is examined, a pixel search is performed in the ruled line candidate area in the vertical or horizontal direction. Is calculated by a numerical value called the degree of scrutiny. Thus, the degree of flaring can be investigated in detail.

According to the eleventh aspect of the present invention, the sharpness is obtained by searching for a pixel in a ruled line candidate area to determine the number of breaks and the length of the break in the ruled line pattern. Add the value converted to
This makes it possible to quantify the degree of shading.

According to the twelfth aspect of the present invention, a processing method for extracting a ruled line or a cell based on the quality or a means for selecting a threshold used in the processing is provided.
A rule line extraction processing method or a threshold value used for the rule line extraction process is selected and set in accordance with a value expressing the image quality by a numerical value. As a result, ruled lines can be extracted with higher accuracy by a processing method or a threshold value adapted to image quality.

According to the thirteenth aspect of the present invention, the means for selecting an image resolution suitable for processing for extracting a ruled line or a cell for each part of the input image,
An image to be used in the subsequent ruled line extraction process is selected based on a value representing the image quality by a numerical value. As a result, when the quality of the image is poor, it is possible to perform processing using a higher quality image, and it is possible to extract ruled lines with higher accuracy.

According to the fourteenth aspect of the present invention, it is possible to calculate both the value for the entire image data and the value for each part of the image data. Thereby, the quality of the image of each part of one image can also be determined.

According to the fifteenth aspect of the present invention, the means for selecting an image resolution suitable for processing for extracting a ruled line or a cell for each part of the input image,
Based on the degree of blurring obtained for each part of the image data, an image having a resolution suitable for each processing of each part of the image data in the subsequent ruled line extraction processing is selected. As a result, it is possible to cope with a partially blurred image.

According to the sixteenth aspect of the present invention, the means for extracting the ruled line or cell uses an image having a resolution different from the image used by the ruled line candidate area extracting means based on the quality of the image. Is also possible. As a result, in the case of a blurred image, the ruled line extraction processing can be performed again using an image having a lower resolution than the image from which the ruled line candidate region has been extracted or the same image, and a more accurate ruled line can be extracted. it can.

According to the seventeenth aspect of the present invention, the means for extracting the ruled line or the cell uses a higher resolution image when it is determined that the image is not blurred. If it is determined that there is, an image with a lower resolution is used. As a result, ruled lines can be extracted with higher accuracy.

According to the eighteenth aspect of the present invention, there is provided a ruled line extracting method for extracting ruled lines from an image including a form separated into a plurality of portions by shading.
A pattern connected in any one of the eight horizontal and diagonal directions is extracted as a connection pattern, and among the extracted connection patterns, those having a certain size are located close to each other. Using the shape of the white pixel area between the patterns, it is determined whether or not the same table is formed.
This makes it possible to accurately extract the table structure of a form separated by shading or the like.

According to the twentieth aspect of the present invention, there is provided a ruled line extracting method for extracting a ruled line from an image including a form separated into a plurality of portions by shading.
A pattern connected in any one of the eight horizontal and diagonal directions is extracted as a connection pattern, and among the extracted connection patterns, those having a certain size are located close to each other. Using the shape of the black pixel at the end of the pattern, it is determined whether the same table is formed. This makes it possible to accurately extract the table structure of a form separated by shading or the like.

According to the twenty-second aspect of the present invention, there is provided a ruled line extracting method for extracting a ruled line from an image including a form separated into a plurality of parts by shading.
A pattern connected in any one of the eight horizontal and diagonal directions is extracted as a connection pattern, and when extracted connection patterns having a certain size are located close to each other, those connection patterns are extracted. Rule lines are extracted from the patterns, and if the correspondences match the rule positions extracted from the respective patterns, it is determined that the same table is to be formed, and the integration process is performed. This makes it possible to accurately extract the table structure of a form separated by shading or the like.

[0032]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a ruled line extraction method according to a first embodiment of the present invention. In FIG. 1, an input image is acquired by an image input unit 101,
The image generating means 102 having different resolutions creates a low-resolution image (FIG. 2A) and a high-resolution image (FIG. 2B) by thinning out data from the input image. Note that a high-resolution image refers to an image having a resolution equal to or lower than the resolution of an input image, and a low-resolution image refers to an image having a lower resolution than a high-resolution image. The ruled line candidate area extracting means 103 extracts ruled line candidate areas using the generated low-resolution image. Image quality evaluation means 104 includes:
The quality of the image is evaluated by searching for pixels in the line candidate area of the high-resolution image.

The image quality evaluation means 104 can evaluate the quality of the entire image data and also can evaluate the quality of each part of the image data. The means 105 for selecting a processing method or a threshold value according to the quality selects a processing method or a threshold value adapted to the image quality based on the result evaluated by the image quality evaluation means 104. The means 106 for selecting an image resolution suitable for each partial process is
An image to be processed is selected based on the image quality.
Alternatively, the resolution of the image used when processing each part of the image is selected based on the image quality obtained for each part of the image. From the above preprocessing, more accurate ruled line extraction processing can be performed by the ruled line extraction processing unit 107.

In the ruled line candidate area extracting means 103, even if it cannot be extracted as a straight line due to the influence of blurring,
Since the low-resolution image is used, the sharpened portion is crushed and can be determined as a straight line, and the accuracy of ruled line candidate region extraction can be improved.

The image quality evaluation means 104 searches for a pixel in the straight line candidate area extracted by the ruled line candidate area extracting means 103, and uses the detected number of breaks and the length of the break to determine the blur of the ruled line. The degree (shearing degree) is calculated. Means 1 for selecting a processing method according to the quality or means for selecting a threshold, using the calculated degree of blur as image quality.
05, it is possible to select a suitable image resolution, a suitable process and a threshold.

According to the present invention, when an image is blurred, a low-resolution image (FIG. 2A) which is less affected by image blurring is selected as a processed image. A feature is that a high-resolution image (FIG. 2B) is selected as a processing image and ruled lines are extracted so that a survey can be performed. As described above, in the present invention, the accuracy of the ruled line extraction process can be improved by performing a quality evaluation of the image as a pre-process of the ruled line extraction process, and then performing a process suitable for the quality based on the evaluation result.

FIG. 3 is a block diagram showing a ruled line extraction method according to a second embodiment of the present invention. As shown in FIG. 3, the ruled line extraction method according to the present embodiment employs a low-resolution image creation unit 3.
01, a high-resolution image creation unit 302, a connection pattern extraction unit 303, a dotted line extraction unit 304, a dotted line image elimination unit 305, a label integration processing unit 306, a ruled line candidate region extraction unit 307, a ruled line attribute determination unit 308, It is composed of a part 309 to be set as a non-target area for calculation processing, a blurring degree calculation unit 310, a ruled line discrimination processing unit 311, a ruled line integration processing unit 312, and a cell extraction unit 313.

In FIG. 3, the input image is a binary image having no extreme inclination. The low-resolution image creation unit 301 and the high-resolution image creation unit 302 reduce the input image using OR thinning. Note that the created low-resolution image and high-resolution image are stored separately from images used for processing described later, and are used for display on a display device. A low-resolution image is used from the connection pattern extraction unit 303 to the ruled line candidate area extraction 307 in FIG.

The connected pattern extracting unit 303 in FIG. 3 labels and extracts patterns connected by eight connections in the low-resolution image obtained in the preprocessing. The pattern “connected by eight connections” refers to a pattern that is connected in any one of eight vertical, horizontal, and diagonal directions. A general technique can be used for the labeling. Each connection pattern in the image obtained by the connection pattern extraction unit 303 is called a partial pattern. Since the circumscribed rectangle of the partial pattern obtained by the labeling is required in the processing described later, the coordinates of the corners of the rectangle obtained by approximating the partial pattern to the rectangle are calculated during the labeling processing.

The dotted line extraction processing unit 304 in FIG. 3 extracts a pattern forming a dotted line from the input image labeled by the connection pattern extraction unit 303. In this case, the pattern constituting the dotted line to be extracted may be a dashed line in which segments of a fixed length are arranged at regular intervals, or a chain line in which segments of different lengths are regularly arranged. There are dotted lines in which size partial patterns are regularly arranged. In this processing, any method may be used as long as a dotted line is extracted. Hereinafter, the dotted line extraction processing unit 304 for extracting a portion where partial patterns of the same size are regularly arranged as a dotted line portion will be described using a table including dotted lines shown in FIG. 4 as an example.

First, as shown in FIG. 5, the circumscribed rectangle 40 of the label calculated by labeling from the pattern of FIG.
A part of the image in 1 is taken out as a strip 501 in the vertical or horizontal direction. Then, the size of the pattern existing inside the strip 501 is compared with the size of the pattern constituting the preset dotted line, and the portion having the same size as the pattern constituting the preset dotted line is compared. A pattern is searched (here, a case where a vertical dotted line is extracted is shown as an example).

Next, as shown in FIG. 6, a portion where a partial pattern having the same size as the previously set size is found is inserted into a strip perpendicular to the strip 501 in FIG. 5 (in this case, in the vertical direction). Cut out as 601. Then, a partial pattern having the same size as the pattern constituting the previously set dotted line is extracted within the range of the strip 601.

Next, as shown in FIG. 7, a pattern in which the extracted partial patterns are arranged at a certain interval and are arranged continuously for a certain number or more is determined as a dotted line part.
This is approximated by a rectangle 701. Then, as shown in FIG. 8, the above processing is repeated to sequentially extract dotted lines, and these are approximated by a rectangle 801.

FIG. 9 is a flowchart showing the operation of the dotted line extraction processing unit 304. In FIG. 9, first, in step S901, a part of the labeled input image is cut out in a strip shape in the horizontal direction or the vertical direction.

Next, in step S902, step S9
It is determined whether or not a partial pattern of a certain size exists within the range cut out in a strip shape at 01. Then, when there is no partial pattern having a certain size, the process is terminated. On the other hand, if it is determined in step S902 that a partial pattern having a certain size exists, the process advances to step S903 to input a strip-shaped image including the extracted partial pattern and perpendicular to the strip used in step S901. Cut out from the image.

Next, in step S904, step S9
It is determined whether or not a partial pattern having a certain size exists inside the strip-shaped image cut out in 03. Then, if there is no portion having a certain size, the process is terminated. On the other hand, if it is determined in step S904 that there is a partial pattern having a certain size, the process proceeds to step S905, where the partial patterns are extracted, and it is determined whether the partial patterns are arranged at regular intervals. If the partial patterns are not arranged at regular intervals, the process ends.

On the other hand, if it is determined in step S905 that the extracted partial patterns are arranged at regular intervals, the flow advances to step S906 to enclose a range in which the partial patterns exist in a rectangle. Then, Step S907
Thus, the range surrounded by the rectangle is defined as the range where the dotted line exists.

The dotted image erasing unit 305 in FIG.
A process of rewriting all black pixels within the range determined to be the range where the dotted line exists in the dotted line extraction process of the preprocessing to white pixels is performed.

In the case of a table separated into a plurality of parts by blurring, the label integration processing unit 306 in FIG. 3 integrates partial patterns to which the same label is assigned, and performs a process of regarding it as one table. Judging only by the positional relationship of the labels, in the two examples as shown in FIG. 10, it is impossible to determine whether or not the lower left cell forms the same table. However, since FIG. 10A originally includes the same table, it is necessary to perform the integration process, and FIG. 10B does not need to perform the integration process because the table is different. In the label integration process 306, it is determined whether or not the two types shown in (a) and (b) of FIG. 10 are regarded as one table, and the label integration process is performed based on the result.

The label integration processing unit 306 determines whether to integrate labels based on the shape of white or black pixels between partial pattern labels, and determines whether or not to integrate labels by using ruled line position information. There are two methods: These two methods are described in turn below.

First, a method for integrating labels by the shape of white or black pixels between labels of partial patterns is described below.
A method of making a determination based on the shape of a white pixel between labels will be described as an example. As shown in FIG. 11 (a), the shape of the white pixel area between the label 1101a and the label 1102a, which originally constitute the same table, has irregularities due to ruled lines as shown by the length of the arrow in the figure. It becomes a shape area. FIG.
As shown in b), the shape of the white pixel area between the label 1101b and the label 1102b, which originally constitute a different table, is a smooth shape without unevenness. The present method utilizes this feature.

The shape unevenness determination uses the length of white pixels between labels. A large overlap between two certain label circumscribed rectangles is extracted as an integration candidate. The change in the length of the white pixel between the candidate labels is investigated. At this time, if there is only a global change, it is possible that the image is tilted (when the difference between the longest value and the shortest value is large even if the variation in the length of the neighboring white pixel is small, 11 (b) is inclined), and a local change is investigated. As a result, if there is a large variation, it is considered that the shape has irregularities. If it is determined that the shape has irregularities, it is integrated with the target label, the size of the circumscribed rectangle is changed, and thereafter the label is regarded as the same.

Similarly, the determination can be made by examining the unevenness of the black pixel area. When investigating the unevenness of the black pixel area, the shape of the end of the label constituting the partial pattern is protruding (in the part where the target label is located in the vertical direction, the vertical ruled line is projected. In the position where the horizontal ruled line is protruding, it can be determined that the same table is configured.

FIG. 12 shows a flowchart of the label integration process when it is determined whether or not to integrate labels based on the shape of white pixels between labels. First, image information and label information are input. In step S1201, two labels (a, b) are selected. If it is determined in step S1202 that the circumscribed rectangles of the label a and the label b do not overlap and the distance between the circumscribed rectangles of the label a and the label b is not smaller than a predetermined value, the process returns to step S1201. Step S
In step 1202, if the circumscribed rectangles of the label a and the label b overlap or the distance between the circumscribed rectangles of the label a and the label b is equal to or smaller than a certain value, the process proceeds to step S1203.
The width of the white pixel between the label a and the label b is checked. In step S1204, it is determined whether the amount of change in the width of the white pixel between the label a and the label b is greater than a predetermined value.
Return to 201. If it is determined in step S1204 that the amount of change in the width of the white pixel between the label a and the label b is greater than a certain value, the process proceeds to step S1205, where the label a and the label b are integrated into a label c. S12
In step 06, the coordinate value indicating the circumscribed rectangle of the label c is calculated, and the process ends.

Next, a method for judging whether or not to integrate labels using the position information of the ruled line will be described. In this method, ruled lines are extracted from two labels, and a determination is made based on the correspondence between the extracted ruled lines. FIG.
As shown in a), first, the ruled lines of the label 1301a and the label 1302a which is the target label are extracted. Then, as shown by a circle in FIG.
When the ruled line 01a is in a protruding state and is located at substantially the same position as the coordinates of the ruled line extracted from the target label 1302a, the originally continuous ruled line is broken and cut off. It can be determined that the same table is configured, and the integration processing can be performed. As shown in FIG. 13 (b), a label 1301b and a label 1302 which originally constitute different tables.
Since the ruled line extracted from b does not have a protruding portion, no integration processing is performed.

FIG. 14 shows a flowchart of a label integration process when it is determined whether or not labels should be integrated using ruled line position information. First, image information and label information are input. In step S1401, two labels a and b
A ruled line is extracted from each. Proceeding to step S1402, if the x coordinate of the ruled line extracted from the label a is not substantially the same as the x coordinate of the ruled line extracted from the label b,
Proceeding to step S1407, the process ends without performing the integration process. In step S1402, the x coordinate of the ruled line extracted from the label a is replaced by the x coordinate of the ruled line extracted from the label b.
If the coordinates are substantially the same, the process proceeds to step S1403. If the y coordinate of the ruled line extracted from the label a is not substantially the same as the y coordinate of the ruled line extracted from the label b, the process proceeds to step S1407, and the integration process is not performed. , And the process ends. If the y-coordinate of the ruled line extracted from the label a in step S1403 is substantially the same as the y-coordinate of the ruled line extracted from the label b, the process advances to step S1404 to protrude from the ruled line extracted from the label a in the direction of the label b. If there is no ruled line, the process advances to step S1407 to terminate the process without performing the integration process. In step S1404, the label a
If there is a ruled line protruding in the direction of the label b among the ruled lines extracted from, the process proceeds to step S1405. If there is no ruled line protruding in the direction of the label a among the ruled lines extracted from the label b, the process proceeds to step S1407. The processing ends without performing the integration processing. If it is determined in step S1405 that there is a ruled line protruding in the direction of the label a among the ruled lines extracted from the label b, the process advances to step S1406 to perform a label integration process and end the process.

The label integration processing unit 306 in FIG. 3 has been described above. Note that the label integration processing unit is effective for more accurately extracting the table structure of a form separated by shading or the like, and separately uses a process corresponding to the above-described label integration process to extract ruled lines in an image of poor quality. It is also possible. Therefore, the present invention is not limited to being incorporated in the processing procedure of the present embodiment as described above.

Next, the ruled line candidate area extraction unit 30 shown in FIG.
7 will be described. The ruled line candidate area extraction unit 307 extracts a ruled line candidate area from the processed image described above. This process may be performed by any method as long as information indicating the area where the image forming the ruled line exists is indicated. Here, an example of the ruled line candidate area extraction processing unit 307 is shown below.

FIG. 15 is a block diagram of the ruled line candidate area extracting unit 307. The ruled line candidate area extracting unit 307 includes a mask processing unit 1501 and a vertical and horizontal ruled line extracting unit 1502. Further, the vertical and horizontal ruled line extraction unit 1502 includes a line segment extraction unit 1503,
It comprises a straight line extraction unit 1504 and a ruled line candidate extraction unit 1505. The description will be given in order according to the flow of the block diagram shown in FIG.

The mask processing unit 1501 performs a mask process described below for each partial pattern obtained by the label integration processing unit 306. In the mask process, the entire image subjected to the label integration process is scanned in two types of masks, a horizontally long one and a vertically long one, and the ratio of the pattern in the mask is calculated. If the calculated value is larger than a predetermined value, the entire mask is regarded as a pattern. On the other hand, if the calculated value is equal to or smaller than the predetermined value, the vertical and horizontal components are extracted by deleting the pattern in the mask. When a plurality of rows or columns continue and the ratio becomes larger than a predetermined value, a rectangular range is created by combining the rows or columns, and the center line of the rectangular range is set as a processing result. Note that in order to prevent a gap from being opened between line segments obtained as a result of the mask processing, it is only necessary to overlap each of the areas where the mask processing is performed. FIG.
FIG. 9 is a diagram illustrating a case where a horizontal component is extracted by using a mask that specifies a rectangular range of 1 × 6 pixels in length and width. In the figure, (
a) shows masks “M1” to “M9” overlapping each other. FIG. 2B shows an example of an image of the original partial pattern, and FIG. 2C shows horizontal components extracted when mask processing is performed using the mask shown in FIG. 1A. .
In the figure, black circles represent extracted horizontal component pixels.

FIG. 17 is a diagram for explaining an embodiment of the processing of the mask processing section 1501 in more detail, and shows a flowchart when the processing is realized by software. FIG.
In step S1701, the entire image subjected to the label integration process is scanned in two types of masks, a horizontally long one and a vertically long one. In step S1702, the ratio of the pattern in the mask, that is, the area of the pattern in the mask with respect to the size of the mask is calculated, and it is determined whether the ratio is greater than a predetermined value. If the decision result in the step S1702 is NO, it is determined in the step S1703 that there is no vertical component or horizontal component, and the process ends. On the other hand, step S1
If the determination result of step 702 is YES, step S170
In step 4, all of the inside of the mask is regarded as a pattern, and all of them are set as vertical or horizontal components. A step S1705 determines whether or not the obtained component is in contact with the upper and lower components or the left and right components. If the decision result in the step S1705 is NO, the component obtained in the step S1706 is output as a processing result.

If the decision result in the step S1705 is YES, the components in contact are integrated in a step S1707, and a rectangular range is created from the integrated components in a step S1708. A step S1709 calculates a center line of the created rectangular range, and a step S1710 outputs the center line as a processing result.

Next, the vertical / horizontal ruled line extracting unit 1502 will be described. For extracting vertical and horizontal ruled line candidates, a low-resolution image equivalent to 100 dpi is used. First, the line segment extraction unit 1503 in the vertical and horizontal ruled line extraction unit 1502 divides the masked pattern into a plurality of parts in the horizontal and vertical directions, and calculates the adjacent projection of the pattern in each of the divided areas in the horizontal and vertical directions. Then, a line segment or a part of a straight line having a certain length is detected by rectangular approximation. Here, the adjacent projection is obtained by adding the projection value of the surrounding row or column to the projection value of the target row or column. The projection value of the target row or column is the sum of the black pixels present in the row or column. By this adjacent projection, a straight line can be accurately detected even when the straight line is inclined and exists over a plurality of rows or a plurality of columns. For this reason, when detecting a large frame such as a block frame, it is possible to detect a straight line constituting the frame even if the frame is inclined.

FIG. 18 is a diagram showing a projection result of a line segment constituting a certain rectangular frame. In FIG. 18, a projection value Ph (i) in a horizontal direction j of a rectangular frame 1801 having a length LY in the vertical direction and a length LX in the horizontal direction is represented by HP (i) and a rectangular frame 1801.
Let VP (j) be the projection value Pv (j) in the vertical direction i of HP (0) = HP (n) = m, HP (1) to HP (1)
(N-1) = 2, VP (0) = VP (m) = n, VP (
1) -VP (n-1) = 2. Thus, the rectangular frame 1
Since the projection value of the portion where the straight line constituting 801 exists is large, the straight line constituting the rectangular frame 1801 can be extracted by calculating the projection value. P (i) = p (ij) + p (ij + 1) + ... + p (i) + ... + p (i + j) (1) FIG. 19 shows the state of the pixel level of the inclined straight line. Is shown. In this case, in FIG. 19, the adjacent projection value is determined as j = 1 in the equation (1). Since only nine black pixels exist in the (i-1) th row, p (i-1) = 9. Since only six black pixels exist in the (i) th row, p (i) =
It becomes 6. Similarly, p (i + 1) = 6. As a result, P (
i) = p (i-1) + p (i) + p (i + 1) = 21

As described above, when the straight line is inclined, and the straight line exists over a plurality of rows or a plurality of columns, the adjacent projection value of the inclined straight line becomes large. Even in a case where the frame is inclined, it is possible to efficiently detect a straight line that forms the frame.

FIG. 20 is a flowchart showing the operation of the adjacent projection process. In FIG. 20, step S200
Reference numeral 1 divides the partial pattern having the same label obtained by the mask processing unit 1501 into a plurality of vertical and horizontal portions. Next, in step S2002, a projection value is calculated within each of the vertical and horizontal division ranges. Step S2003
Then, the surrounding projection values are added to the calculated projection values. Further, in step S2004, an adjacent projection value is calculated based on the above equation (1). The line segment extraction unit 1503
Based on adjacent projection values of the partial pattern with respect to the mask processing image, a line segment or a part of a straight line having a certain length in the horizontal direction and the vertical direction is detected by rectangle approximation.

That is, when the ratio between the adjacent projection value of the divided partial pattern and the vertical / horizontal division length is equal to or greater than a predetermined threshold, a rectangular range obtained by combining a plurality of continuous rows or columns is represented by a straight line. Is the position where the candidate exists. Note that a line segment of a certain length or a part of a straight line detected by this rectangular approximation is called a “rectangular line segment”.

For example, as shown in FIG. 21A, the oblique horizontal line 2101 is divided into three parts, and the position where the adjacent projection value is equal to or larger than the threshold value is calculated. As a result, as shown in FIG. 21B, three rectangular line segments 2102 divided in the horizontal direction can be obtained.

Here, when a straight line 2101 that is obliquely inclined is detected by using a normal projection method instead of the adjacent projection method, the projection value becomes small.
It becomes impossible to detect 1. On the other hand, when the number of divisions of the partial pattern is increased and the division length is shortened in order to detect the oblique straight line 2101 by the normal projection method, a large number of short straight lines constituting the character are also detected, and the character and the frame are detected. It becomes difficult to distinguish between

On the other hand, by using the adjacent projection method, even when the straight line 2101 is obliquely inclined, it is possible to increase the adjacent projection value without extremely reducing the division length of the partial pattern. Therefore, a relatively long straight line forming the frame can be accurately detected while being distinguished from a short straight line forming the character.

FIG. 22 is a flowchart showing the operation of the line segment extraction unit 1503. In FIG. 22, first, in step S2201, it is determined whether the ratio between the adjacent projection value of the divided partial pattern and the vertical and horizontal division lengths is equal to or greater than a predetermined threshold. If it is determined that the ratio between the adjacent projection value and the vertical / horizontal division length is not equal to or greater than the predetermined threshold, the process proceeds to step S2203, and it is determined that no line segment exists. On the other hand, if it is determined in step S2201 that the ratio between the adjacent projection value and the vertical / horizontal division length is equal to or greater than the predetermined threshold, the process advances to step S2202, and it is determined that a line segment exists.

Next, in step S2204, if it is determined that the pattern regarded as a line segment in step S2202 is not in contact with the lines above and below it, the process proceeds to step S2205, where a rectangular line segment is set. On the other hand, if it is determined in step S2206 that the pattern regarded as a line segment in step S2202 is in contact with a line segment existing above and below it, the process proceeds to step S2206, and the pattern regarded as a line segment and the pattern above and below it are determined. Integrate with existing line segments. Then, in step S2207, the line segments integrated in step S2206 are detected as rectangular line segments.

Returning to the description of FIG.
No. 4 detects long straight lines by integrating neighboring line segments among the rectangular approximated line segments extracted in the line segment extraction. The detected straight line performs rectangular approximation.

The straight line extracting unit 1504 includes a line segment integrating unit and a straight line detecting unit. If the unbroken rectangular line segments are in contact or connected with each other as in the line segments y and z shown in FIG. 23A, the line segment integrating unit integrates these rectangular line segments to form a long straight line. I do. Furthermore, even if the rectangular line segments are not connected to each other as in the case of the line segments x and y shown in FIG. 23 (a), if the vertical distance is within the number j of rows or columns added during the adjacent projection, These rectangular line segments are integrated into a long straight line (FIG. 23B). If the ratio of the length of the straight line integrated by the line integration unit to the length of the rectangle approximating the partial pattern having the same label is equal to or greater than a predetermined threshold, the straight line detection unit Is determined to be a candidate.

FIG. 24 is a diagram for explaining the processing of the line segment integrating section of the straight line extracting section 1504 in more detail, and shows a flowchart in the case where the processing is realized by software. FIG.
In step S2401, a distance between the detected rectangular line segments is calculated. In step S2402, it is determined whether or not the calculated distance is within the number of rows or columns to be added during the adjacent projection. If the determination result in step S2402 is NO
If so, it is assumed that the integration of the rectangular lines is not performed in step S2404. On the other hand, if the decision result in the step S2402 is YES, a step S2403 carries out integration of rectangular lines.

As described above, the ruled line candidate area extraction unit 307
The line segment extracting unit 1503 and the straight line extracting unit 1504 have been described. Finally, as the processing of the ruled line candidate area extracting unit 307, the ruled line candidate extracting unit 1505 shown in FIG.
At 504, the extracted straight line is extracted as a ruled line candidate, and the process is terminated.

When the above processing for the low-resolution image is completed, the processing using the high-resolution image proceeds. Referring back to FIG. 3, the ruled line attribute determining unit 308 examines the ruled line candidate area obtained by processing the low resolution image in high resolution in detail, and determines the ruled line attribute (dotted line or solid line).
Is determined.

The ruled line attribute determination unit 308 extracts the attribute using the low resolution image in the dotted line extraction processing unit 304, and the dotted line may be connected. Plays a role in preventing accidental attachment. The processing method is as follows:
In the same manner as described above, the determination is made based on whether or not labels of the same size are regularly arranged using the high-resolution image labeled by the connection pattern extraction unit 303 ′.

Next, in the non-target area setting unit 309 of the shading degree calculation processing in FIG. 3, the area extracted as the dotted line area by the dotted line extraction processing unit 304 and the area obtained as the dotted line area by the ruled line attribute determination unit 308 Is set as a region not to be subjected to the flaring degree calculation process.

Then, the degree of sharpness calculation unit 310 shown in FIG. 3 calculates the degree of the quality of the ruled line of the image being processed as
Calculate the degree of scrutiny (degree of scrutiny). Based on the result, an image, a threshold value, and the like to be used in the subsequent processing are determined.

The processing range is limited within the region for calculating the degree of blurring set in the preprocessing (that is, processing is performed only on the straight line candidates), and the degree of blurring of the target image is calculated. A high resolution is used for the resolution of the image to be used so that the degree of blur can be accurately investigated.

When the pattern is separated into more patterns, it is determined that the degree of blur is high. Further, it can be determined that the degree of blur is higher as the black pixels disappear and the number of white pixels increases. In order to obtain the degree of blurring based on this determination criterion, the number of pattern breaks and the length of pattern breaks (white pixels) are calculated in the pixel search processing within the straight line candidate area. The processing is performed on both the vertical and horizontal ruled line candidate areas, and the calculated result is converted into per unit length. The sum of the number of pattern breaks per unit length and the pattern break length is defined as the degree of blur. Assuming that the number of searched pixels in the straight line candidate area is L, the number of white pixels in the area is a, and the number of white pixel areas in the area is b, the degree of blurring c can be expressed as in equation (2).

C = a / L + b / L (2) FIG. 25 shows a specific example of calculating the degree of blur. FIG. 25 (a) to (
In d), the number L of search pixels in the ruled line candidate area approximated by a rectangle is 20 pixels. In FIG. 9A, the number of white pixels a is 6 and the number of white pixel regions b is 6, so that the degree of blurring c = 6/2.
0 + 6/20 = 0.6. Also, in FIG. 3B, since there are two white pixels a and two white pixel areas, the blurring degree c
= 2/20 + 2/20 = 0.2. From this,
It can be seen that the greater the number of places in the white pixel area, the higher the degree of blur. Comparing the calculated degrees of sharpness with FIGS. 7A and 7B, it can be said that the degree of sharpness and the degree of sharpness of the actual image match.

In FIG. 9C, the number of white pixels a is three and the white pixel area b is one, so that the degree of bleeding c = 3/20 + ／.
0 = 0.2. In FIG. 11D, since the number of white pixels is 7 and there is one white pixel area, the degree of blur c = 7/20
+ 1/20 = 0.4. From this, it can be seen that the greater the number of white pixels, the higher the degree of blur. Also, FIG.
Comparing the calculated degree of sharpness with (d), it can be said that the degree of sharpness matches the degree of sharpness of the actual image.

The following processing method and parameters are set in accordance with the degree of shading obtained as a result. To explain the outline of the processing after this, when the degree of blurring is low, the allowable blurring length is shortened, and the image is examined in detail with a high-resolution image to determine whether or not the ruled line candidate area is a ruled line. judge
(Ruled line determination processing unit 311 'in FIG. 3). If the degree of sharpening is high, the allowable length of sharpening is set relatively long, and processing is performed on a low-resolution image with little influence of sharpening (ruled line determination processing unit 311 in FIG. 3). Processing for complementing the blurring of the ruled line is added (the blurring complementing processing unit 312 in FIG. 3).

In calculating the degree of sharpness, the entire image is not represented by one degree of sharpness, but a plurality of degrees of sharpness, such as a degree of sharpness with respect to a horizontal ruled line and a degree of sharpness with respect to a vertical ruled line, or a finer portion. The degree of rubbing may be calculated.
In this case, when referring to an image having a different resolution for a certain part in the image in a later process, parameter setting and the like are also performed for each part for which the degree of blur is set.

Further, as a result of calculating the degree of blur, if it is determined that the image is a blurred image, the ruled line extraction process should be re-extracted using a lower resolution image in which the influence of the blur is reduced. By doing so, it is possible to achieve even higher precision. Conversely, when there is no blurring, even if a high-resolution image is used, it is not necessary to consider the adverse effects of the blurring, so it is possible to process with a higher resolution image and investigate the image in detail It can be.

When the degree of blur is calculated and the quality of the target image is determined, the image to be processed differs depending on the degree of blur as shown in FIG. If the degree of sharpness is high, that is, if it is determined that there is a sharpening, the low-resolution image is used for the subsequent processing.If the degree of sharpness is low, that is, it is determined that there is no sharpening, the high-resolution image is used for the subsequent processing. If the sharpness is calculated to be 0, it is determined that the image is crushed, and the conventional ruled line extraction processing 31 using the input image as it is is performed.
Perform 4. In FIG. 3, the ruled line discrimination processing unit 311 after the sharpness is determined to be high and the ruled line discrimination processing unit 311 ′ after the blurring is determined to be low differ only in the image used and have the same processing content. is there. Ruled line determination processing unit 311,
311 'determines whether or not the area is a ruled line by checking the degree of unevenness of the ruled line candidate area. Therefore, when the value of the degree of blur is high, that is, when an image of poor quality is used, the degree of unevenness is calculated to be high due to the effect of straight lines and discontinuities or lines becoming thinner due to fading and are originally ruled lines. However, it may not be determined as a ruled line. In this case, the effect of blurring can be suppressed by performing ruled line discrimination processing using an image with a lower resolution. When the degree of shading is low, it is considered that there is almost no influence on the shape of the ruled line. Therefore, by performing the ruled line discrimination process using a high-resolution image, it is possible to make a more detailed determination.

Although the image used is different, the ruled line discrimination processing unit 3
11 and 311 'perform the same processing of determining whether the pattern is a ruled line or a part erroneously extracted from a character based on the degree of unevenness of the ruled line candidate pattern. Those with a high degree of unevenness are determined to be erroneous extractions from characters, and those with a low degree of unevenness are determined to be ruled lines.

The ruled line discrimination processing sections 311 and 311 'are used to accurately detect the horizontal or vertical straight line extracted as a ruled line candidate area in the case of a horizontal line and the upper and lower ends in the case of a vertical line. A search for a pattern in the ruled line candidate area is performed. FIG. 26 is a diagram illustrating a search result of a pattern in a ruled line candidate area. In FIG. 26, characters of “research department” are written between the ruled lines, and the character string of “research department” is in contact with the lower ruled line. The upper ruled line is extracted as a ruled line candidate 2601, and the lower ruled line is extracted as a ruled line candidate 2603. Since the characters "ken" and "ku" are close to each other, "
Are extracted as ruled line candidates 2602.

The ruled line discrimination processing units 311, 311 'search for patterns in these ruled line candidates 2601, 2602, 2603. At this time, the number of changes in the search direction is counted. Here, since the pattern of the ruled line candidate 2601 is originally a ruled line, it is searched straight, and the value of the number of changes in the search direction becomes small. On the other hand, ruled line candidates 2602
Is a character, the search path is bent reflecting the shape of the character, and the value of the number of changes in the search direction increases.

As a result, a pattern having a small value of the number of changes in the search direction is regarded as a ruled line, and a pattern having a large value of the number of changes in the search direction is not regarded as a ruled line, whereby the pattern of the ruled line candidate 2601 is determined to be a ruled line. In addition to this, it is possible to determine that the pattern of the ruled line candidate 2602 is not a ruled line.

Also, only when there is no adjacent pixel in the search direction, a diagonal pixel is searched, and as long as there is an adjacent pixel in the search direction, the search is performed in the same direction. Ruled line candidate 260 touching the character string
For pattern 3 as well, the value of the number of changes in the search direction becomes smaller, so that it is possible to determine that the pattern of ruled line candidate 2603 is a ruled line.

The ruled line determination processing units 311 and 311 'determine whether the pattern is a ruled line based on whether the number of changes in the search direction, that is, whether the degree of unevenness of the pattern exceeds a set threshold value. 27 shows an example of setting the threshold. FIG.
In the above, the character string of "research section" is entered above the shaved ruled line, and the blurred ruled line is extracted as a ruled line candidate 2702, and the characters "ken", "ku" and "part"
Are close to each other, a part of the character string of “research section” is extracted as a ruled line candidate 2701.

The ruled line candidate 2701 is erroneously extracted from a part of the character string of “research section”.
In the vicinity of 01, there is a remaining pattern of the character string of “research section” extracted as the ruled line candidate 2701. Therefore, since another pattern exists near the ruled line candidate 2701, the threshold value TH1 of the degree of unevenness of the pattern of the ruled line candidate 2701 is set to be small. As a result, ruled line candidate 2
It is possible to more accurately exclude a part of the character string of “research section” extracted as 701 from ruled line candidates.

On the other hand, since the ruled line candidate 2702 is correctly extracted from the ruled line and is apart from the character string of “research section”, no other pattern such as a character string exists near the ruled line candidate 2702. . Therefore, ruled line candidate 27
The threshold value TH2 of the unevenness degree of the pattern No. 02 is set to be large. This makes it possible to more accurately determine a ruled line candidate 2702, which has been sharpened and thus has an increased degree of unevenness, without being excluded from the ruled line candidates.

FIGS. 28 and 29 are flowcharts showing the processing for calculating the degree of unevenness. Note that this flowchart shows a case where the search is performed in the horizontal direction. FIG.
8, in step S2801, a rectangular area of a ruled line candidate is set as a search range. Next, step S280
In step 2, the horizontal coordinates of the thinnest portion of the pattern in the rectangular area of the ruled line candidate are calculated, and the center point of the pattern at the horizontal coordinates of the thinnest portion is calculated. Then, the center point of this pattern is set as the search start point. Here, the start point of the search is set to the narrowest part of the pattern because the narrowest part is likely to be a ruled line, and the search for a straight line serving as a frame can be performed more reliably.
Next, in step S2803, the search direction of the straight line is set to the right. Next, in step S2804, step S28
The search start point set in 02 is set as a target pixel.

Next, in step S2805, the initial value of the count K for counting the length of the blank area is set to 0. In step S2806, a variable hup for counting the number of times the search direction has changed upward or downward to the left is set to 0, and in step S2807, a variable hbl for counting the number of times the search direction has changed downward or to the left downward is set to 0.
Set to. Also, in step S2808, a variable hver for counting the number of times of crossing the vertical line is set to 0.

FIG. 30 is a diagram showing an arrangement relationship of pixels according to the second embodiment of the present invention. In FIG. 30, D1 is a pixel adjacent to the upper left diagonally adjacent, D2 is a pixel adjacent to the left, D3 is a pixel adjacent diagonally to the lower left, D4 is a pixel adjacent to the upper, D5 is a lower adjacent pixel, D6 is an upper right diagonally adjacent pixel, D7 is a right adjacent pixel, and D8 is a lower right diagonally adjacent pixel.

When the search proceeds from the pixel of interest to the pixel D1, the search changes to the upper left. When the search advances to the pixel D3 from the pixel of interest, the search direction changes to the lower left. When the search progresses to D6, the search direction changes to the upper right, and when the search advances from the target pixel to the pixel D8, the search direction changes to the lower right.

Returning to the description of FIG. 28, step S280
In step 9, it is determined whether the target pixel is inside the rectangular area of the pattern to be searched. If the target pixel is not inside the rectangular area of the pattern to be searched, the process advances to step S2823 to count the length of the blank area. After setting the value of the variable K to be performed to 0, the process proceeds to step S2824.

On the other hand, if it is determined in step S2809 that the pixel of interest is inside the rectangular area of the pattern to be searched, the process advances to step S2810, and the process proceeds to step S2810.
It is determined whether there is a black pixel at the position of D7 described with reference to 0. If the determination is YES, the process proceeds to step S2811,
Let the pixel of interest be the pixel at position D7,
If there is a black pixel at the position, the search proceeds to the right.

On the other hand, if the determination in step S2810 is NO, the flow advances to step S2812 to set D6 for the target pixel.
It is determined whether there is a black pixel at the position of. Judgment is YE
In the case of S, the process proceeds to step S2813 to increase the value of the variable hup by 1, and then proceeds to step S2814.
The pixel of interest is the pixel at the position D6. Thus, the number of times the search direction changes to the right as well as the search direction changes to the right is counted. After the target pixel has moved to the position of D6, steps S2809 to S28
Step 11 is repeated.

On the other hand, if the determination in step S2812 is NO, the flow advances to step S2815 to set D8 for the target pixel.
It is determined whether there is a black pixel at the position of. Judgment is YE
In the case of S, the process proceeds to step S2816 to increase the value of the variable hbl by 1, and then proceeds to step S2817.
The pixel of interest is the pixel at the position D8. Thus, the number of times that the search direction has changed to the lower right while the search direction changes to the lower right is counted. After the target pixel moves to the position of D8, steps S2809 to S28
Step 11 is repeated.

On the other hand, if the determination in step S2815 is NO, the process advances to step S2818 to determine whether the variable K for counting the length of the blank area is equal to or smaller than a threshold. If the result of the determination is YES, the flow advances to step S2819 to determine whether the pixel of interest is a black pixel and both the pixels above and below (D4 and D5) are black pixels. If the determination in step S2819 is YES, it is assumed that the vehicle crosses the vertical line, and the flow advances to step S2820 to set the variable h
After incrementing ver by 1, the process advances to step S2821 to set the target pixel as the pixel at the position D7. On the other hand, if the determination in step S2819 is NO, step S282
Proceeding directly to 1, the target pixel is set as the pixel at the position of D7. In step S2822, the variable K for counting the length of the blank area is increased by 1, and the flow returns to step S2809.

On the other hand, if the determination in step S2818 is NO, the process advances to step S2824 to determine whether the search direction is set to the right. Here, if the search direction is set to the right, the process proceeds to step S2825, where the search direction is set to the left, and then the process proceeds to step S2826, where the search start point determined in step S2802 is set as the target pixel. Then, the process returns to step S2809 to search leftward.

On the other hand, if it is determined in step S2824 that the search direction is not right, the flow advances to step S2827 to subtract the value corresponding to the image inclination from the values of the variables hup and hbl, thereby obtaining the image inclination. Is corrected. The value corresponding to the image inclination is calculated in advance from the average value of the inclination of the ruled line candidate. Then, step S282
At 8, the values of the variables hup, hbl, and hver are converted to values per a certain length. Next, in step S2829, when the value of the variable hup, the variable hbl, or the variable hver converted to a value per a certain length is equal to or larger than the threshold, the pattern is excluded from ruled line candidates.

The variables hup and h obtained by the above procedure
The value of bl and the variable hver is set as the degree of unevenness, and the ruled line determination processing unit determines whether or not the line is a ruled line using the threshold value of the degree of unevenness described with reference to FIG. In addition, variable hup, variable h
The value of bl and the value of the variable hver may be added, and the result may be converted into a ratio to the length of the straight line.

In the flowcharts of FIGS. 28 and 29, the search in the horizontal direction has been described. However, the search can be performed in the vertical direction by changing the search direction from left to right and up and down. It is.

As described above, the ruled line discrimination processing section 31 in FIG.
The processing of 1, 311 'has been described. Next, the ruled line integration processing unit 312 in FIG. 3 will be described below. As shown in FIG. 3, the processing by the ruled line integration processing unit 312 is performed after the processing by the above-described ruled line determination processing unit 311 when the blurring degree calculation unit 310 determines that the image is blurred. The ruled line integration processing unit 312 performs a process of integrating ruled lines separated by blurring.

Although not shown in FIG. 3, even when it is determined that there is no blur, the ruled line integration processing unit 312 'is inserted between the ruled line integration determination processing unit 311' and the cell extraction unit 313 ', and It is also possible to further improve the extraction accuracy.
In this case, the ruled line integration processing unit 312 sets a large threshold value indicating the length between ruled lines that can be integrated when it is determined that the image is blurred, whereas the ruled line integration processing unit 312 after determining that the image is not blurred. In ', the threshold value indicating the length between the ruled lines that can be integrated is set to a small value or set to 0 so that no processing is performed.

For example, as shown in FIG. 31A, horizontal lines 3101 to 3106 have been extracted up to the pre-processing. Here, the horizontal line 3102 and the horizontal line 3103 are separated by the influence of blurring. In this case, they are on the same Y coordinate and X
It is checked whether the distance in the direction is equal to or less than a predetermined value. When the horizontal lines 3102 and 3103 satisfy this condition, FIG.
As shown in FIG. 1 (b), the horizontal line 3102 and the horizontal line 3103 are integrated into a horizontal line 3112. As a result, the shading 310
The horizontal lines 3111 to 1115 obtained by correcting 7 are obtained.

In the case of a vertical line, they are on the same X coordinate,
In addition, straight lines whose distance in the Y direction is equal to or smaller than a predetermined value are integrated. If the target frame is an irregular frame, the lengths and positions of the straight lines vary, so that integration of straight lines that are extremely far apart is not performed.

FIG. 32 is a flowchart showing the operation of the process when the ruled line integration processing unit 312 integrates horizontal straight lines. In FIG. 32, first, in step S3201
Then, two horizontal straight lines extracted by the ruled line candidate area extraction processing are extracted two by two. Next, in step S3202, it is determined whether the distance in the X direction between the two straight lines extracted in step S3201 is equal to or less than a predetermined threshold. If the result of the determination is NO, the process ends.

On the other hand, the determination in step S3202 is YES
In step S3203, the process proceeds to step S3203.
It is determined whether the distance in the Y direction between the two horizontal straight lines extracted in step 1 is equal to or smaller than a predetermined threshold. If the result of the determination is NO, the process ends.

On the other hand, the determination in step S3203 is YES
In the case of, the process proceeds to step S3204, and step S320
The two horizontal straight lines extracted in 1 are integrated and regarded as one straight line.

FIG. 33 is a flow chart showing the operation of processing when the ruled line integration processing unit 312 integrates vertical straight lines. FIG. 33 is almost the same as the case where the horizontal straight lines are integrated.

In FIG. 33, first, at step S3301
Then, two vertical straight lines extracted by the ruled line candidate area extraction processing are extracted two by two. Next, in step S3302, it is determined whether the distance in the Y direction between the two straight lines extracted in step S3301 is equal to or smaller than a threshold. If the result of the determination is NO, the process ends. On the other hand, if the determination in step S3302 is YES, the process advances to step S3303 to determine whether the distance in the X direction between the two vertical straight lines extracted in step S3301 is equal to or less than a predetermined threshold. If the result of the determination is NO, the process ends. On the other hand, step S33
If the determination in step 03 is YES, the process advances to step S3304, and the two vertical straight lines extracted in step S3301 are integrated and regarded as one straight line.

Next, a straight line constituting the frame is determined from the ruled lines obtained by the above processing, and a rectangular range in which the upper, lower, left, and right sides are surrounded by the frame is extracted as a cell. The cell extraction units 313 and 313 ′ in FIG. 3 play this role. Note that a method for extracting cells in the cell extraction units 313 and 313 ′ is as follows.
Any method may be used. Here, the following method will be described as an example.

FIG. 34 is a view for explaining an example of the cell extraction processing sections 313 and 313 'according to the second embodiment of the present invention. In FIG. 34A, a straight line forming the form 3401 is determined by obtaining ruled line candidates from the input image. Next, as shown in FIG.
The form 3401 is divided into rows by determining the horizontal frame from the straight line constituting 1. Next, as shown in FIG. 34C, a cell is extracted from the form 3401 by determining a vertical frame from a straight line constituting the form 3401.
Here, it is checked whether a nested structure 3402 exists in the cells extracted from the form 3401. Next, FIG.
As shown in d), the nested structure 3402 is regarded as a new form, and cells are extracted from the nested structure 3402.

FIG. 35 is a diagram for explaining a method of determining the above-described horizontal frame. In FIG. 35, a form 340
For 1, the horizontal frame and the vertical lines (I) to (VI) are extracted, and the horizontal line is determined as a group one by one in order from the top row to determine a horizontal frame. For example,
The horizontal straight line and the horizontal straight line are determined as a pair of horizontal lines, and the horizontal straight line and the horizontal straight line are set as a horizontal frame. This horizontal frame determination method is performed according to the following procedure.

1) From the horizontal straight lines determined to be ruled line candidates in the process up to the cell extraction processing unit, two horizontal straight lines are extracted in order from the top, and these are extracted as a set of two horizontal straight lines. Make it a candidate. 2) If the length of the two horizontal straight lines as the set candidates is the same or the lower straight line is longer, the two straight lines are regarded as a set of horizontal straight lines. At this time, if the lengths of the two horizontal straight lines are different, the longer horizontal straight line can be reused.

3) If the lower horizontal straight line among the two horizontal lines as a set candidate is shorter, the lower horizontal straight line is further set as a horizontal frame candidate and the lengths of these horizontal straight lines are compared. If the lower horizontal line is shorter in the right or left direction, the lower horizontal line is set as a candidate for a set, and the lengths of these horizontal lines are compared.

4) As a result of comparing the lengths of the horizontal straight lines, 2)
If it is found that there is no horizontal line that satisfies the horizontal line that satisfies the above condition, only in this case, the upper horizontal line and the lower horizontal line are combined.

5) After the processing of the bottom horizontal line, if there is an unprocessed horizontal line at the top, the unprocessed horizontal line and the reusable horizontal line are again used in order from the top again. 1) ~
Perform the processing of 4) to create a set of two horizontal straight lines.

FIG. 36 is a diagram for explaining a method of determining the above-described vertical frame. Referring to FIG.
With respect to 1, horizontal straight lines and vertical straight lines (I) to (VI) are extracted, and vertical straight lines (I) to (VI) reaching both upper and lower sides of the set horizontal frame are determined as vertical frames. For example, since the vertical straight line (I) and the vertical straight line (VI) both reach the horizontal straight line and the horizontal straight line regarded as a horizontal frame, the vertical straight line (VI)
I) and a vertical straight line (VI) are a vertical frame.

FIG. 37 is a diagram for explaining the above-described cell extraction method. In FIG. 37, the horizontal straight line and the vertical straight lines (I) to (VI) are extracted from the form 3401, and a rectangular area surrounded by horizontal and vertical frames on all four sides is extracted as a cell. For example, a rectangular area surrounded by a horizontal straight line and a vertical straight line (IV), a horizontal straight line and a vertical straight line (I) is extracted as a cell, and a horizontal straight line and a vertical straight line (VI), a horizontal straight line and a vertical straight line (IV) are extracted. ) Is extracted as a cell.

FIG. 38 is a diagram for explaining a method of extracting the nested structure described above. In FIG. 38, when a rectangular area surrounded by a horizontal straight line and a vertical straight line (III) and a horizontal straight line and a vertical straight line (I) is extracted as a cell, the cell has a nested structure 3402. Therefore, the nested structure 34
The horizontal frame and the vertical frame are determined from the straight lines that constitute 02, and a rectangular area surrounded by the horizontal frame and the vertical frame on all four sides is extracted as a cell in the nested structure 3402. For example, horizontal straight line and vertical straight line
(III), a rectangular area surrounded by a horizontal straight line and a vertical straight line (I), a horizontal straight line and a vertical straight line (II), a horizontal straight line and a vertical straight line (II).
I) and a rectangular area surrounded by horizontal and vertical lines (I
II), a rectangular area surrounded by a horizontal straight line and a vertical straight line (II) is extracted as a cell.

FIG. 39 shows a flowchart of the horizontal frame determining method shown in FIG. 35. In FIG. 39, first,
In step S3901, the top two horizontal straight lines are extracted from the straight lines determined to be ruled line candidates. Next, in step S3902, of the two straight lines taken out, the upper horizontal straight line is upline, the lower horizontal straight line is blline,
And In step S3903, it is determined whether the straight line upline and the straight line blline overlap in the vertical direction. If the determination result is NO, the process advances to step S3904 to replace the straight line below the straight line blline with a new straight line blll.
ine, and this processing is performed by using a straight line upline and a straight line bll.
The process is continued until it is determined that there is a vertical overlap with the ine.

On the other hand, if the decision result in the step S3903 is Y
In the case of ES, the process proceeds to step S3905, where
It is determined whether the length of ne and the straight line blline are the same, or whether the straight line blline is longer than the straight line upline. If there are a straight line upline and a straight line blline that satisfy the above condition, the process advances to step S3908 to proceed to step S3908.
Let lline be a set of two horizontal straight lines (horizontal frame of two sets). The fact that the length of the straight line upline is the same as the length of the straight line blline indicates a state as shown in FIG.

On the other hand, if it is determined in step S3905 that the length of the straight line upline is different from the length of the straight line blline, and that the straight line blline is shorter than the straight line upline, the flow advances to step S3906 to proceed to step S3906.
a straight line blline which is equal to or longer than the straight line upline below ne
It is determined whether or not 2 exists. If the result of the determination is YES, the flow advances to step S3907 to change the straight line blline to the straight line blline2, and the flow advances to step S3908. It should be noted that the fact that the straight line blline is shorter than the straight line upline indicates a state as shown in FIG. 40 (b), and that the straight line blline is longer than the straight line upline.
A state as shown in FIG. 40 (c) is shown.

If the decision result in the step S3906 is NO, the process advances to a step S3908 to set the straight line upline and the straight line blline as a set of two horizontal frames. Next, in step S3909, vertical frame determination processing and nesting processing shown in the flowchart of FIG. 41 are performed. Next, step S391
0, a straight line upli defined as two frames in step S3908
It is determined whether the length of ne and the line blline are different. When the result of the determination is YES, the flow proceeds to step S3911, where a straight line that is long in the right direction or a straight line that is long in the left direction is reusable, and the flow proceeds to step S3912. The straight line extending rightward means the straight line blline shown in FIG. 40 (d), and the straight line extending leftward is the straight line blline shown in FIG. 40 (e).

On the other hand, if the decision result in step S3910 is N
In the case of O, the process proceeds to step S3912, where the straight line bllin
It is determined whether a straight line exists below e. If the determination is NO, the flow advances to step S3913 to extract the straight line blline and the straight line below it, and returns to step S3902.

If the decision result in the step S3912 is YES, the process advances to a step S3914 to remove the processed straight line from the processing target except for the reusable straight line. Next, in step S3915, it is determined whether an unprocessed straight line exists. If the result of the determination is NO, the process ends.
On the other hand, if the decision result in the step S3915 is YES,
Proceeding to step S3916, two straight lines are extracted from the uppermost portion, and the process returns to step S3902.

FIG. 41 is a flowchart showing details of the vertical frame determination and nesting processing in step S3909. In FIG. 41, first, in step S4101, a horizontal set of two sets in the top row is selected, and in step S4102, both the upper end and the lower end of the vertical straight line of the ruled line candidate are set in the horizontal set of two sets. Select the one that has reached the frame. Next, step S
In 4103, the selected vertical straight line is determined as a vertical frame. next,
In step S4104, a rectangular area surrounded by a horizontal set of two sets and a vertical frame of two sets is extracted. Then, step S4
In 105, it is determined whether or not there is a horizontal straight line reaching both ends of the left and right vertical frames of the rectangular area. Judgment result is N
In the case of O, the process of selecting the horizontal frame of the next row is performed, and the same process as the above-described steps S4102 to S4107 is executed on the newly determined horizontal frame of the double set.

On the other hand, if the decision result in the step S4105 is Y
In the case of ES, the flow advances to step S4106 to regard the rectangular area as a new small table (nested rectangle), and determine the horizontal frame of the small rectangular area existing in the nested rectangular area. Then, as shown in step S4107, a process of determining a vertical frame in the nested rectangular area is performed.

As described above, the ruled line extraction method of the second embodiment shown in FIG. 3 has been described. According to the present invention, ruled lines can be extracted with higher accuracy regardless of the quality of an image. FIG. 4 shows an example of a hardware configuration for realizing the ruled line extraction method described in the first embodiment and the second embodiment of the present invention.
It is shown in FIG. Image data such as a form is input to a processing device 420 such as a computer via an input device 4201 such as a scanner.
2 Here, a series of processing by the above-described ruled line extraction method is performed. The processing result in the middle can be output to an output device 4203 such as a display, and the extraction processing can be performed while confirming the progress of the processing based on the output image. In FIG. 42, the processing device is shown as a single unit. However, a processing device including a plurality of computers or a plurality of processing devices (computers) connected via a network
In this case, each processing means of the above-described ruled line extraction method is distributed to each processing device.

The present invention relates to a program relating to a ruled line extraction method capable of extracting ruled lines with higher accuracy regardless of the quality of an image in the processing device 4202. FIG. 43 provides a software program according to the present invention. It is a figure explaining a method. The program is provided by any of the following three methods, for example.

(A) Processing device 4301 such as computer
Installed and provided. In this case, the program or the like is pre-installed before shipment, for example. This corresponds to the above described method of distributing the software program of the ruled line extraction method.

(B) Provided by being stored in a portable storage medium. In this case, the programs and the like stored in the portable storage medium 4302 are stored in a processing device 430 such as a computer.
1 storage device.

(C) Provided from a server on the network 4303. In this case, basically, a processing device 4301 such as a computer downloads a program or the like stored in the server 4304 to acquire the program or the like.

[0142]

As described above in detail, according to the present invention, the quality of an image is evaluated by calculating the degree of blurring of the image, and based on the calculated image quality, the image used for the ruled line extraction processing is converted into a low-resolution image and a high-resolution image. By selecting from among the resolution images, it is possible to correctly extract a ruled line that has not been correctly extracted in the past due to blurring, thereby improving the accuracy of ruled line extraction.

Further, according to one aspect of the present invention, the degree of blur, which indicates the degree of blur, is obtained by searching pixels in the meridian candidate area in the vertical or horizontal direction, and determining the number of pattern breaks per ruled line unit length and the number of breaks It is calculated by calculating the length. This makes it possible to digitize the degree of sharpness, and to perform a ruled line extraction processing method according to the digitized degree of sharpness,
A threshold value used for the ruled line extraction processing, an image used, and the like can be set.

Further, according to one aspect of the present invention, the degree of blurring can be determined by using either a value for the entire image data or a value for each part of the image data. As a result, the quality of the image of each part of one image can be determined, and as a result, an image having a resolution suitable for processing of each part of the subsequent image data can be selected.

Further, according to one aspect of the present invention, a ruled line extracting method for extracting ruled lines from an image including a form separated into a plurality of parts by shading is described with reference to the vertical, horizontal, horizontal
A pattern connected in any one of the eight diagonal directions is extracted as a connection pattern, and when the connection patterns having a certain size are located close to each other, the following three A determination is made as to whether or not to integrate the same table using one of the methods. 1) Judgment is made using the shape of the white pixel region between those patterns. 2) Judgment is made using the shape of the black pixel area at the end of those patterns. 3) Judgment is made using the correspondence between the extracted ruled line positions. This makes it possible to more accurately extract the table structure of a form separated by shading or the like.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a ruled line extraction method according to a first embodiment of the present invention.

FIG. 2A is a diagram illustrating an example of a low-resolution image.
(B) is a diagram showing an example of a high-resolution image.

FIG. 3 is a block diagram showing a ruled line extraction method according to a second embodiment of the present invention.

FIG. 4 is a diagram (part 1) illustrating a dotted line extraction process of a ruled line extraction method according to a second embodiment of the present invention.

FIG. 5 is a diagram (part 2) for explaining the dotted line extraction processing of the ruled line extraction method according to the second embodiment of the present invention.

FIG. 6 is a diagram (part 3) for explaining the dotted line extraction processing of the ruled line extraction method according to the second embodiment of the present invention.

FIG. 7 is a diagram (part 4) illustrating a dotted line extraction process of the ruled line extraction method according to the second embodiment of the present invention.

FIG. 8 is a diagram (part 5) illustrating a dotted line extraction process of the ruled line extraction method according to the second embodiment of the present invention.

FIG. 9 is a flowchart showing an operation of a dotted line extraction processing unit of a ruled line extraction method according to a second embodiment of the present invention.

FIG. 10 is a diagram illustrating an example of data input to a label integration processing unit of a ruled line extraction system according to a second embodiment of the present invention. FIG. An example of separation is shown, and (b) shows an example in which two tables are originally configured.

FIG. 11 is a diagram showing an example of a white pixel area between labels;
(A) is a case where the same table is originally constituted,
(B) shows an example in which a different table is originally configured.

FIG. 12 is a diagram illustrating a flowchart of a label integration process when determining whether to integrate labels based on the shape of white (or black) pixels between labels.

13A and 13B are diagrams illustrating an example of a correspondence relationship between ruled lines of a target label. FIG. 13A illustrates an example in which the same table is originally configured, and FIG. 13B illustrates an example in which a different table is originally configured. It is shown.

FIG. 14 is a diagram illustrating a flowchart of a label integration process when it is determined whether or not to integrate labels by using ruled line position information (correspondence relationship).

FIG. 15 is a diagram illustrating an example of a block diagram of a ruled line candidate area extraction unit according to a second embodiment of the present invention.

FIG. 16 is a diagram showing a mask processing method according to a second embodiment of the present invention.

FIG. 17 is a flowchart illustrating an operation of a mask process according to the second embodiment of the present invention.

FIG. 18 is a diagram showing a projection result of line segments constituting a frame.

FIG. 19 is a diagram illustrating adjacent projection processing in a line segment extraction unit according to the second embodiment of the present invention.

FIG. 20 is a flowchart illustrating an operation of an adjacent projection process in a line segment extraction unit according to the second embodiment of the present invention.

FIG. 21 is a diagram showing a line segment extraction process according to the second embodiment of the present invention.

FIG. 22 is a flowchart illustrating an operation of a line segment extraction processing unit according to the second embodiment of the present invention.

FIG. 23 is a diagram showing a straight line extraction process according to the second embodiment of the present invention.

FIG. 24 is a flowchart illustrating an operation of line segment integration processing in a straight line extraction processing unit according to the second embodiment of the present invention.

FIG. 25 is a diagram illustrating a skewness calculating process according to the second embodiment of the present invention.

FIG. 26 is a diagram showing a result of a pattern search of a ruled line candidate area in a ruled line determination process according to the second embodiment of the present invention.

FIG. 27 is a diagram illustrating a setting example of a threshold value of the degree of unevenness of the pattern of the ruled line candidate area in the ruled line determination processing unit according to the second embodiment of the present invention.

FIG. 28 is a flowchart illustrating a roughness calculating process in a ruled line discrimination processing unit according to the second embodiment of the present invention.

FIG. 29 is a continuation of the flowchart in FIG. 28;

FIG. 30 is a diagram showing an arrangement relationship of pixels according to a second embodiment of the present invention.

FIG. 31 is a view for explaining horizontal straight line integration processing in a ruled line integration processing unit according to the second embodiment of the present invention.

FIG. 32 is a flowchart illustrating an operation of a process when a horizontal straight line is integrated in a ruled line integration processing unit according to the second embodiment of the present invention.

FIG. 33 is a flowchart showing an operation of a process for integrating vertical straight lines in a ruled line integration processing unit according to the second embodiment of the present invention.

FIG. 34 is a diagram illustrating processing of a cell extraction processing unit according to the second embodiment of the present invention.

FIG. 35 is a diagram illustrating a method for determining a horizontal frame in a cell extraction processing unit according to the second embodiment of the present invention.

FIG. 36 is a diagram illustrating a vertical frame determination method in the cell extraction processing unit according to the second embodiment of the present invention.

FIG. 37 is a diagram illustrating a cell extraction method in a cell extraction processing unit according to the second embodiment of the present invention.

FIG. 38 is a diagram for explaining a method of extracting a nested structure in the cell extracting unit according to the second embodiment of the present invention.

FIG. 39 is a flowchart illustrating a method for determining a horizontal frame in a cell extracting unit according to the second embodiment of the present invention.

FIG. 40 is a diagram illustrating an arrangement of horizontal straight lines according to the second embodiment of the present invention.

FIG. 41 is a flowchart showing vertical frame determination and nesting processing in a cell extraction unit according to the second embodiment of the present invention.

FIG. 42 is a diagram showing an example of a hardware configuration for executing a ruled line extraction method according to the present invention.

FIG. 43 is a diagram illustrating a method for providing a software program and the like according to the present invention.

FIG. 44 is a diagram showing an example of a conventional method in form processing that does not require frame information in advance.

[Explanation of symbols]

 101 image input means 102 image generation means with different resolutions 103 ruled line candidate area extraction means 104 image quality evaluation means 105 means for selecting a processing method or threshold value according to quality 106 means for selecting an image resolution suitable for each partial processing 107 Ruled line extraction processing (cell extraction) means 301 Low-resolution image creation unit 302 High-resolution image creation unit 303, 303 'Connected pattern extraction unit 304 Dotted line extraction processing unit 305 Dotted line image deletion unit 306 Label integration processing unit 307 Ruled line candidate area extraction unit 308 Ruled line attribute determining unit 309 Non-target area setting unit for blurredness calculation processing 310 Blurredness calculating unit 311, 311 'Ruled line discriminating processing unit 312 Ruled line integration processing unit (blurring complementing process) 313 cell extraction unit 314 Conventional processing 1501 Mask processing unit 1502 vertical and horizontal ruled line extraction 1503 line segment extraction unit 1504 line extraction unit 505 ruled line candidate extraction unit 4201 input device 4202 processor (computer) 4203 Output device 4301 processing unit (computer) 4302 a storage medium 4303 network 4304 server

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5B009 PA03 PA04 PB03 5B029 CC27 DD05 EE12 EE19 5C076 AA19 AA22 5L096 EA03 EA37 FA03 FA18 FA38 FA54 FA64 FA70 GA10 GA15 GA26 GA28 GA34 GA51

Claims (23)

[Claims] 1. A ruled line extraction method for extracting ruled lines forming a form in image data, comprising: means for generating images of different resolutions from an input image; means for evaluating the quality of the input image; A ruled line extraction method, comprising: means for selecting an image having a resolution corresponding to the quality from images having a resolution; and means for extracting a ruled line or a cell from the selected image. 2. A ruled line extraction method for extracting ruled lines constituting a form in image data, comprising: means for generating images having different resolutions from an input image; and means for generating images having different resolutions from the input image. Means for evaluating the quality of the input image using the generated image; and a processing method for extracting a ruled line or a cell based on the quality evaluated by the means for evaluating the quality of the input image, or a method for performing the processing. Means for selecting a threshold value to be selected; means for selecting an image resolution suitable for processing for extracting a ruled line or a cell for each part of the input image; and selecting the ruled line from the image of the selected image resolution. Alternatively, a book based on a processing method or a threshold value selected by a processing method for extracting cells or a threshold value used for the processing is selected. Ruled line extraction method, characterized in that it comprises, means for extracting a ruled line or cell constituting. 3. The image processing apparatus according to claim 1, wherein said means for generating images having different resolutions generates a low-resolution image having a large reduction ratio and a high-resolution image having a small reduction ratio when reducing an input image. 2. The ruled line extraction method described in 2. 4. The means for evaluating the quality of the input image,
3. The ruled line extraction method according to claim 2, wherein the quality is expressed by a numerical value. 5. The ruled line extraction method according to claim 2, further comprising a ruled line candidate area extracting means for extracting a ruled line candidate area in advance before performing processing by the means for evaluating the quality of the input image. 6. The ruled line extraction method according to claim 5, wherein the ruled line candidate area extracting means performs a ruled line candidate extraction process using a low resolution image. 7. The image processing apparatus according to claim 1, further comprising: means for extracting a dotted line using a low-resolution image and removing a pattern forming the dotted line before performing the processing by the ruled line candidate area extracting means. The ruled line extraction method according to claim 6. 8. A ruled line attribute determining means for determining a straight line attribute by performing a dotted line extracting process and a ruled line extracting process on the ruled line candidate area extracted by the ruled line candidate area extracting means using a high-resolution image. 7. The ruled line extraction method according to claim 6, further comprising: 9. The image quality evaluation device according to claim 5, wherein the image quality evaluation unit evaluates the image quality using an image having a different resolution from the image used in the ruled line candidate area extraction unit. Ruled line extraction method. 10. The image quality evaluation means, when examining the ruled line candidate area, performs a pixel search in the ruled line candidate area in the vertical or horizontal direction, and the blurring degree of the image is referred to as the blurring degree. 6. The ruled line extraction method according to claim 5, wherein the calculated degree is used as a quality of an image. 11. The blurring degree is obtained by searching for a pixel in a ruled line candidate area to obtain the number of breaks and the length of the break in the ruled line pattern, and adding the converted values per unit length. 11. The ruled line extraction method according to claim 10, wherein: 12. A processing method for extracting a ruled line or a cell based on the quality or a means for selecting a threshold used in the processing, the ruled line extraction processing corresponding to a value representing the image quality by a numerical value. 5. The method according to claim 4, wherein a threshold value used in a ruled line extraction process is selected and set.
Ruled line extraction method described. 13. A means for selecting an image resolution suitable for a process for extracting a ruled line or a cell for each part of the input image, comprising: 5. The ruled line extraction method according to claim 4, wherein an image used for the extraction processing is selected. 14. The ruled line extraction method according to claim 10, wherein the degree of blurring can be calculated for both a value for the entire image data and a value for each part of the image data. 15. A means for selecting an image resolution suitable for processing for extracting a ruled line or a cell for each part of the input image, wherein the means for selecting the image resolution is based on the degree of blurring determined for each part of the image data. 15. The ruled line extraction method according to claim 14, wherein in the subsequent ruled line extraction process, an image having a resolution suitable for each processing of each part of the image data is selected. 16. The image processing apparatus according to claim 16, wherein the means for extracting the ruled line or the cell can use an image having a different resolution from the image used by the ruled line candidate area extracting means based on the quality of the image. The ruled line extraction method according to claim 5. 17. A means for extracting a ruled line or a cell uses a higher-resolution image when it is determined that the image is not blurred, and uses a higher-resolution image when it is determined that the image is blurred. 17. The ruled line extraction method according to claim 16, wherein a low-resolution image is used. 18. A ruled line extraction method for extracting a ruled line from an image including a form separated into a plurality of parts by shading, wherein the ruled line is adjacent to a target pixel in any of eight vertical, horizontal, and oblique directions in an input image. The connected patterns are extracted as a connected pattern by searching for the pixels in sequence, and when the extracted connected patterns having a certain size are located close to each other, those patterns are detected. A ruled line extraction method for extracting a ruled line forming a form after performing an integration process by determining that the same table is configured using the shape of a white pixel region between the two. 19. The ruled line extraction method according to claim 18, wherein if there is unevenness in the shape of the white pixel area between the patterns, it is determined that the same table is formed, and the integration process is performed. A ruled line extraction method characterized by determining that a different table is formed if there is no unevenness in the shape of a region and performing an integration process. 20. A ruled line extracting method for extracting a ruled line from an image including a form separated into a plurality of portions by shading, wherein the ruled line is adjacent to a target pixel in any of eight vertical, horizontal, and oblique directions in an input image. The connected patterns are extracted as a connected pattern by sequentially searching for the pixels that exist, and when those having a certain size are located close to each other among the extracted connected patterns, those patterns are detected. A ruled line extraction method for extracting the ruled lines forming the form after determining that the same table is configured using the shape of the black pixel located at the end of the form and performing the integration processing. 21. The ruled line extraction method according to claim 20, wherein when the shape of the black pixel area at the end of the pattern is a protruding shape, it is determined that the same table is formed and the integration processing is performed. Characteristic ruled line extraction method. 22. A ruled line extraction method for extracting a ruled line from an image including a form separated into a plurality of parts by shading, wherein the input image is adjacent to a target pixel in any of eight vertical, horizontal, and oblique directions. The connected patterns are extracted as a connected pattern by sequentially searching for the pixels that exist, and when those having a certain size are located close to each other among the extracted connected patterns, those patterns are detected. A ruled line extraction method for extracting a ruled line from each pattern, and determining that the same table is formed if the correspondence is the same with the ruled line position extracted from each pattern, and performing an integration process. 23. A storage medium storing a program for extracting a ruled line forming a form in image data, the method comprising: generating images of different resolutions from an input image; and generating images of different resolutions from the input image. Evaluating the quality of the input image using the image generated by the step of performing, and a processing method for extracting ruled lines or cells based on the quality evaluated in the step of evaluating the quality of the input image, or Selecting a threshold used for processing; selecting an image resolution suitable for processing for extracting a ruled line or a cell from each part of the input image; and A processing method for extracting a ruled line or a cell or a threshold value used for the processing is selected by means for selecting. Has been processing method or computer-readable storage medium storing a program for executing the steps, the process comprising the processing apparatus for extracting ruled lines or cells constituting a form based on a threshold.