JP4753889B2 - Line extraction processing apparatus, line extraction processing program, recording medium storing line extraction processing program, and line extraction processing method - Google Patents

Description

  The present invention relates to a technique for extracting a line such as a character or an edge of a line in multi-value image data.

  In recent years, with the widespread use of digital cameras and scanners, the use of various image processing technologies including image compression technology for multi-value image data, OCR (Optical Character Recognition) technology, and pre-processing for printing has increased.

  Many of these techniques improve processing performance by using different processes in accordance with the characteristics of the image for each area, instead of processing the entire image singly.

  For example, in the OCR, erroneous recognition is prevented by detecting a character area in an image and executing recognition processing only on the detected character area. In image compression, there is a technique for preferentially allocating capacity over other areas to a character area where readability is required or a line area where deterioration is conspicuous.

  From the above, technology for analyzing image features for each region plays a very important role. Among them, extraction of line drawings such as characters is particularly important.

  In general, a Sobel filter is often used for line drawing extraction. However, although the Sobel filter can measure the edge strength, it does not know whether the pixel value is changing in the increasing direction or conversely decreasing. For this reason, it is difficult to specify whether the edge is around the line or the other edge.

Patent Document 1 proposes a method for determining a line drawing. In this method, an input image is differentiated, + P is marked for a pixel whose differential value is positive and whose absolute value is greater than or equal to a threshold value, and −P Mark If the positions of + P and -P are close to each other, it is determined that the line drawing is a line drawing, thereby enabling the line drawing to be extracted.
JP 2000-163598 A

  However, the technique disclosed in Patent Document 1 tends to detect short lines and fine points that are not visually recognized as lines at locations where the pixel value changes greatly (for example, halftone dots).

  The technique disclosed in Patent Document 1 is relatively effective for detecting a line drawing having a strong edge. However, line extraction in image data such as image data captured by a digital camera or scanned image data of a printed material, Or if you want to extract lines with weak edges, such as thin letters, you must lower the threshold value. For this reason, the marks of P and -P increase.

  Since such P and -P marks have only information that the sign is reversed, it is not possible to extract lines of various densities with this information alone. For this reason, even an inappropriate part such as a photograph is determined to be a line, and the detection accuracy falls. In some cases, a line drawing having a specific edge strength cannot be detected.

  In addition, in the technique disclosed in Patent Document 1, P and −P marks are increased for image data such as photographic image data and scanned image data of a printed matter, and lines and lines to inappropriate places are displayed. Judgment and detection accuracy will drop. Also, there are cases where a line drawing having a specific edge strength cannot be detected.

  Furthermore, since the technique disclosed in Patent Document 1 performs line extraction analysis in only one direction, the length of the line is not taken into consideration at all, and halftone dots and dot-like noise are not considered. Tends to be extracted as a line.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide a line extraction processing apparatus that performs line extraction efficiently.

According to one aspect of the present invention, there is provided a line extraction processing unit which extracts data corresponding to a line included in the image picture data, an image input unit for inputting image data, a first direction image data First direction differential processing means for generating first direction differential value data and first direction characteristic differential value data having a part of the first direction differential value data as a first direction characteristic differential value. The first direction characteristic differential value data generation processing means for generating and the characteristic pixel satisfying a predetermined condition among the characteristic pixels taking the first direction characteristic differential value are in a direction perpendicular to the first direction. First direction line extraction processing means for extracting as pixels representing line data in the second direction.

  Preferably, the first direction characteristic differential value data generation processing unit sets, as the first direction characteristic differential value, a first direction differential value having a maximum absolute value in a predetermined section of the first direction differential value data.

  Preferably, the first direction line extraction processing unit has a first direction feature differential value of the second feature pixel existing in the first predetermined range in the first direction of the first feature pixel as the first feature pixel. The difference between the absolute value of the first direction feature differential value of the first feature pixel and the absolute value of the first direction feature differential value of the second feature pixel is the opposite of the first direction feature differential value of the feature pixel. In response to the determination that the value is equal to or less than the first predetermined value, the first feature pixel and the second feature pixel are extracted as pixels constituting the edge of the line in the second direction.

  Preferably, the first direction line extraction processing unit has a first direction feature differential value of the second feature pixel existing in the first predetermined range in the first direction of the first feature pixel as the first feature pixel. The difference between the absolute value of the first direction feature differential value of the first feature pixel and the absolute value of the first direction feature differential value of the second feature pixel is the opposite of the first direction feature differential value of the feature pixel. In response to the determination that the value is less than or equal to the first predetermined value, the first feature pixel, the second feature pixel, and the first feature pixel and the second feature pixel in the first direction The existing pixels are extracted as pixels constituting a line in the second direction.

  Preferably, it further includes first direction feature differential value data noise removal processing means for removing noise from the first direction feature differential value data.

  Preferably, the first direction characteristic differential value data noise removal processing means has a difference between the first characteristic pixel and the first characteristic pixel that is equal to or less than a second predetermined value. The number of connected feature pixels for the first feature pixel is calculated using the third feature pixel adjacent in the direction of the feature pixel as the feature pixel connected to the first feature pixel, and the number of connections is equal to or less than a third predetermined value. The first direction feature differential value data of the feature pixel connected to the first feature pixel and the first feature pixel is removed as noise from the first direction feature differential value data.

According to another aspect of the present invention, there is provided a line extraction processing unit which extracts data corresponding to a line included in the image picture data, an image input unit for inputting image data, a first direction image data First direction differential processing means for generating first direction differential value data and first direction characteristic differential value data having a part of the first direction differential value data as a first direction characteristic differential value. The first direction characteristic differential value data generation processing means for generating and the characteristic pixel satisfying a predetermined condition among the characteristic pixels taking the first direction characteristic differential value are in a direction perpendicular to the first direction. First direction line extraction processing means for extracting pixels representing line data in the second direction, and first direction for differentiating the image data in the second direction to generate second direction differential value data. Two-direction differential processing means and second-direction differential value A second direction feature differential value data generation processing unit for generating second direction feature differential value data in which a part of the data is a second direction feature differential value, and a feature pixel that takes the second direction feature differential value , Second direction line extraction processing means, first direction line extraction processing means, and second direction line extraction processing for extracting feature pixels satisfying a predetermined condition as pixels representing line data in the first direction. Each direction line data synthesizing means for creating line data by synthesizing the pixels extracted by the means.

  Preferably, each direction feature differential value data synthesis processing unit for synthesizing the first direction feature differential value data and the second direction feature differential value data to generate synthesized feature differential value data is further provided.

  Preferably, first direction feature differential value data noise removal processing means for removing noise from the first direction feature differential value data, and second direction feature differential value for removing noise from the second direction feature differential value data. At least one of data noise removal processing means is provided.

  Preferably, the first direction characteristic differential value data generation processing unit sets, as the first direction characteristic differential value, the first direction differential value having the maximum absolute value in the predetermined section of the first direction differential value data, and the second direction The feature differential value data generation processing unit sets the second direction differential value having the maximum absolute value in the predetermined interval in the second direction differential value data as the second direction characteristic differential value.

  Preferably, the first direction line extraction processing unit has a first direction feature differential value of the second feature pixel existing in the first predetermined range in the first direction of the first feature pixel as the first feature pixel. The difference between the absolute value of the first direction feature differential value of the first feature pixel and the absolute value of the first direction feature differential value of the second feature pixel is the opposite of the first direction feature differential value of the feature pixel. In response to the determination that the value is equal to or less than the first predetermined value, the first feature pixel and the second feature pixel are extracted as pixels constituting the edge of the line in the second direction, and the second direction line extraction is performed. The processing means is configured such that the second direction feature differential value of the third feature pixel existing within the second predetermined range in the second direction of the first feature pixel is the second direction feature differential value of the first feature pixel. The value is opposite in sign and the absolute value of the second direction feature differential value of the first feature pixel and the second direction feature derivative of the third feature pixel The first feature pixel and the third feature pixel are extracted as the pixels constituting the edge of the line in the first direction in response to the determination that the difference from the absolute value of is not more than the second predetermined value. To do.

  Preferably, the first direction line extraction processing unit has a first direction feature differential value of the second feature pixel existing in the first predetermined range in the first direction of the first feature pixel as the first feature pixel. The difference between the absolute value of the first direction feature differential value of the first feature pixel and the absolute value of the first direction feature differential value of the second feature pixel is the opposite of the first direction feature differential value of the feature pixel. In response to the determination that the value is less than or equal to the first predetermined value, the first feature pixel, the second feature pixel, and the first feature pixel and the second feature pixel in the first direction The existing pixels are extracted as pixels constituting a line in the second direction, and the second direction line extraction processing means exists within the second predetermined range in the second direction of the first feature pixel. The second feature differential value of the third feature pixel is opposite in polarity to the second direction feature differential value of the first feature pixel. In response to the determination that the difference between the absolute value of the second direction feature differential value of the second feature differential value and the absolute value of the second direction feature differential value of the third feature pixel is equal to or less than the second predetermined value. The pixel, the third feature pixel, and the pixel existing between the first feature pixel and the third feature pixel in the first direction are extracted as pixels constituting a line in the first direction.

  Preferably, the first direction line extraction processing means extracts the pixel representing the line data from the feature pixels taking the combined feature differential value instead of the first direction feature differential value, and the second direction line extraction processing means includes: Instead of the second direction feature differential value, a pixel representing line data is extracted from the feature pixel taking the composite feature differential value.

  Preferably, the first direction feature differential value data noise removal processing means has a difference between the first feature pixel and the first feature pixel that is equal to or less than a third predetermined value. The number of connected feature pixels for the first feature pixel is calculated as a feature pixel connected to the first feature pixel as a feature pixel that is adjacent to the first feature pixel, and the number of connections is equal to or less than a fourth predetermined value. The first direction feature differential value data of the first feature pixel and the feature pixel connected to the first feature pixel in the second direction is removed as noise from the first direction feature differential value data. The first direction feature differential value data is used as the first direction feature differential value data noise removal processing means. The first direction feature differential value data noise removal processing means has a first difference between the first feature pixel and the second direction feature differential value equal to or less than a fifth predetermined value. A feature pixel adjacent to the feature pixel in the first direction, As a feature pixel to be connected to one feature pixel, the number of connected feature pixels for the first feature pixel is calculated, and the second direction feature is determined according to the determination that the number of connections is equal to or less than a sixth predetermined value. The second direction feature differential value data is removed from the differential value data by removing the first feature pixel and the second direction feature differential value of the feature pixel connected to the first feature pixel in the first direction.

Step According to yet another aspect of the present invention, the computer having an arithmetic unit, a program for executing a process of extracting data corresponding to a line included in the image picture data, the arithmetic unit inputs image data And a step of differentiating the image data in the first direction to generate first direction differential value data, and a calculation unit converting a part of the first direction differential value data to the first direction characteristic differential value. And generating a first direction feature differential value data, and among the feature pixels having the first direction feature differential value, the calculation unit selects a feature pixel satisfying a predetermined condition perpendicular to the first direction. And extracting as pixels representing line data in a second direction which is a direction.

  According to still another aspect of the present invention, a computer-readable recording medium storing the line extraction program is provided.

Step According to yet another aspect of the present invention, the computer having an arithmetic unit, a program for executing a process of extracting data corresponding to a line included in the image picture data, the arithmetic unit inputs image data And a step of differentiating the image data in the first direction to generate first direction differential value data, and a calculation unit converting a part of the first direction differential value data to the first direction characteristic differential value. And generating a first direction feature differential value data, and among the feature pixels having the first direction feature differential value, the calculation unit selects a feature pixel satisfying a predetermined condition perpendicular to the first direction. A step of extracting pixels representing line data in a second direction which is a direction, a step of differentiating the image data in the second direction to generate second direction differential value data, and a calculation Part is second A step of generating second direction characteristic differential value data in which a part of the differential value data is set as a second direction characteristic differential value, and a calculation unit satisfying a predetermined condition among characteristic pixels having the second direction characteristic differential value A step of extracting feature pixels as pixels representing line data in the first direction, and an arithmetic unit combining pixels in the first direction and the second direction extracted as pixels representing line data; To create line data.

  According to still another aspect of the present invention, a computer-readable recording medium storing the line extraction program is provided.

According to yet another aspect of the present invention, there is provided a method of extracting data corresponding to a line included in the image picture data, comprising the steps of: inputting image data, by differentiating the image data in a first direction, the A step of creating one-direction differential value data, a step of generating first direction characteristic differential value data in which a part of the first direction differential value data is a first direction characteristic differential value, and a first direction characteristic differential value are taken. Extracting a feature pixel satisfying a predetermined condition from the feature pixels as a pixel representing line data in a second direction perpendicular to the first direction.

According to yet another aspect of the present invention, there is provided a method of extracting data corresponding to a line included in the image picture data, comprising the steps of: inputting image data, by differentiating the image data in a first direction, the A step of creating one-direction differential value data, a step of generating first direction characteristic differential value data in which a part of the first direction differential value data is a first direction characteristic differential value, and a first direction characteristic differential value are taken. Extracting a feature pixel satisfying a predetermined condition from the feature pixels as a pixel representing line data in a second direction which is perpendicular to the first direction; Differentiating in the direction to generate second direction differential value data; generating second direction characteristic differential value data in which part of the second direction differential value data is the second direction characteristic differential value; Two-way characteristic differential value A feature pixel satisfying a predetermined condition is extracted as a pixel representing line data in the first direction, and the first direction and the first extracted as pixels representing line data are extracted. Creating line data by synthesizing pixels in two directions.

  According to the present invention, line extraction can be performed efficiently.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same parts are denoted by the same reference numerals. Their names and functions are also the same. Therefore, detailed description thereof will not be repeated.

[Embodiment 1]
The line extraction processing apparatus according to the first embodiment extracts a line component in a direction orthogonal to the direction by analyzing the arrangement of pixel values (gradation values) in a certain direction.

  In the following, a process for analyzing the arrangement of pixel values in the X direction and extracting a line component in the Y direction in the XY coordinate system will be described. The X direction is assumed to be the horizontal direction and the Y direction is assumed to be the vertical direction. However, the X direction may be any direction as long as the Y direction is orthogonal to the X direction.

  FIG. 1 is a diagram illustrating a specific example of the hardware configuration of a computer 1 on which the line extraction processing device 202 according to the present embodiment is mounted, and is a block diagram illustrating a specific example of the hardware configuration of a general computer. .

  Referring to FIG. 1, a computer 1 includes a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 103, a RAM (Random Access Memory) 105, and a hard disk drive (HDD: Hard Disk) that control the entire apparatus. Drive 107, a reading unit 109 that reads information from the recording medium 2 such as a flexible disk drive (FDD: Flexible Disk Drive) or a CD-ROM (Compact Disk-Read Only Memory) drive, and an input constituted by a keyboard, a mouse, and the like Unit 111, communication unit 113 for communicating with the outside, and display unit 115.

  The line extraction processing apparatus 202 according to the present embodiment is configured by hardware of the computer 1 and software executed by the CPU 101 stored in a storage unit such as the ROM 103, and a function for performing line extraction processing described later is as follows. The CPU 101 is mainly formed in the CPU 101 by reading and executing a program stored in a storage unit such as the ROM 103. Or at least one part of the said function may be comprised by the hardware of the computer 1 shown by FIG.

  FIG. 2 is a block diagram showing a specific example of a functional configuration for functioning as the line extraction processing device 202 in the computer 1 according to the present embodiment and performing line extraction processing.

  With reference to FIG. 2, the functional configuration of the line extraction processing apparatus 202 according to the first embodiment will be described.

  The function of the line extraction processing device 202 includes an input image data buffer 204, an X-direction differentiation processing unit 205, an X-direction differential value data buffer 206, an X-direction feature differential value data generation processing unit 207, and an X-direction feature differentiation. A value data buffer 208, a Y-direction line extraction processing unit 209, and a Y-direction line data buffer 210 are included.

  These components can be configured mainly by a dedicated LSI (Large-Scale Integration) that realizes the function of the line extraction processing device and a memory. However, it may be configured by the CPU 101 of the computer 1 or other hardware. Note that such a dedicated LSI functions as an X-direction differentiation processing unit 205, an X-direction feature differential value data generation processing unit 207, and a Y-direction line extraction processing unit 209.

  The input image data buffer 204, the X-direction differential value data buffer 206, the X-direction characteristic differential value data buffer 208, and the Y-direction line data buffer 210 are realized by the ROM 103, the RAM 105, the HDD 107, or the like.

Below, the said component of the line extraction processing apparatus 202 is demonstrated.
The input image data buffer 204 stores multi-valued image data input from the image input device 201.

  The X direction differentiation processing unit 205 reads image data from the input image data buffer 204. Then, differential processing is performed in the X direction, and the differential value data in the X direction is given to the X direction differential value data buffer 206. Note that the X-direction differentiation processing performed by the X-direction differentiation processing unit 205 may be anything as long as it is a process that can determine the inclination of the pixel value and the rate of change of the pixels in the X direction.

FIG. 3 is a diagram illustrating a filter for calculating a difference from an adjacent pixel.
As illustrated in FIGS. 3A to 3D, the X-direction differentiation processing unit 205 may simply use a filter that calculates a difference from an adjacent pixel. In the present embodiment, for example, a filter shown in FIG.

  Returning to FIG. 2, the X-direction differential value data buffer 206 stores the X-direction differential value data given from the X-direction differential processing unit 205.

  The X direction characteristic differential value data generation processing unit 207 reads X direction differential value data from the X direction differential value data buffer 206. Then, the X-direction feature differential value data created by extracting the X-direction feature differential value is given to the X-direction feature differential value data buffer 208.

  FIG. 4 is a diagram for explaining the X-direction characteristic differential value. 4A is a diagram showing input multi-valued image data, FIG. 4B is a diagram showing an arrangement of pixel values along a broken line 401 in the X direction, and FIG. 4C is a diagram of FIG. It is a figure which shows the X direction differential value data produced | generated by the differential process of a X direction with respect to the arrangement | sequence of the pixel value shown by 4 (b).

The X-direction characteristic differential value data will be described with reference to FIG.
Here, the “characteristic differential value” refers to a differential value having a maximum absolute value within a section in which the differential value is positive or negative in the differential value data. For example, in FIG. 4C, differential values indicated by points 402 to 405 are characteristic differential values.

Further, “feature pixel” refers to a pixel that takes a characteristic differential value.
The “characteristic differential value data” is data including a characteristic differential value extracted from the differential value data. In the characteristic differential value data, the differential value other than the characteristic differential value in the differential value data is expressed by a value that can be distinguished from the characteristic differential value. For example, a differential value other than the characteristic differential value is expressed by a predetermined numerical value, or a predetermined interval in which the differential value other than the characteristic differential value is obtained is expressed by a numerical value different from the characteristic differential value. Such a value can be freely set as long as it can be distinguished from the characteristic differential value. Moreover, you may make it hold | maintain the information which shows whether each differential value is a characteristic differential value as characteristic differential value data. In the present embodiment, for example, a differential value other than the characteristic differential value is expressed by zero, which is a differential value that cannot be the characteristic differential value. FIG. 5 is based on the differential value data shown in FIG. It is a figure which shows the produced characteristic differential value data. In FIG. 5, the x-axis direction indicates the pixel position, and the y-axis direction indicates the differential value.

  Like the columnar graph shown in FIG. 5, the characteristic differential value data is expressed by a characteristic differential value (corresponding to points 402 to 405) and a differential value (zero) other than the characteristic differential value.

  FIG. 6 is a diagram illustrating another example of differential value data and characteristic differential value data. FIG. 6A shows differential value data, and FIG. 6B is characteristic differential value data created based on the differential value data shown in FIG. In FIG. 6, the x-axis direction indicates the pixel position, and the y-axis direction indicates the differential value.

  As shown in FIG. 6A, the differential value data is divided into positive or negative continuous sections 70, 72, 74, and 76. In each section, the differential value having the maximum absolute value indicated by points 71, 73, 75, 77 is the characteristic differential value.

  Like the columnar graph shown in FIG. 6B, the characteristic differential value data includes the characteristic differential values (points 71, 73, 75, 77) in FIG. 6A and differential values other than the characteristic differential values (zero). It is expressed by

  Thus, in this embodiment, the differential value data is separated into the characteristic differential value and the differential value other than the characteristic differential value, and the differential value other than the characteristic differential value is set to zero.

  In the X direction differential value data of photographic image data or image data scanned by a scanner, one edge is expressed by many different differential values. Generally, in one edge expressed by many different differential values, the differential value having the maximum absolute value determines the strength of the apparent edge. For this reason, by performing the processing as described above, it is possible to improve the extraction accuracy in the Y-direction line extraction processing described later and to increase the processing speed.

  In addition, if the length of the positive or negative continuous section is increased, the amount of memory is required. However, by defining the maximum size of the section length and dividing the section length exceeding the maximum size into one or more sections with the maximum section length and the remaining sections, Approximate processing can be realized even in a system with a limited amount of memory.

  In addition, when it is desired to extract a line of edge strength within a predetermined range, a restriction may be placed on a value range that can be taken by the characteristic differential value, and a characteristic differential value that deviates from the value range may be excluded from the characteristic differential value.

  For example, a feature differential value that takes a value of d_min_x or more and d_max_x or less may not be extracted from the differential value data as a feature differential value.

  d_min_x and d_max_x can be freely set. For example, d_min_x is set to 30, d_max_x is set to 50, and the like.

  As will be apparent from the subsequent processing, the line determination is determined by the value and position of the characteristic differential value, and therefore the portion that is not the characteristic differential value does not affect the line determination.

  Returning to FIG. 2, the X-direction feature differential value data buffer 208 stores the X-direction feature differential value data provided from the X-direction feature differential value data generation processing unit 207.

  The Y-direction line extraction processing unit 209 reads the X-direction feature differential value data from the X-direction feature differential value data buffer 208 and outputs Y-direction line data.

Here, the Y direction line data will be described.
The Y direction line extraction processing unit 209 analyzes the arrangement of the X direction characteristic differential value data in the X direction. Then, for the feature pixel b existing within the threshold thick_max_x pixels in the + X direction or the −X direction of the feature pixel a taking a feature derivative value A other than 0, the feature differential value B of the feature pixel b is positive or negative. On the other hand, if the difference between the absolute value | A | of the characteristic differential value A and the absolute value | B | of the characteristic differential value B is within the threshold value_max_x, the characteristic pixels a and b are determined as the edges of the line. Then, the Y direction line data is set to a value indicating that the feature pixels a and b are line information. The value indicating the line information may be arbitrary, but 1 is used here.

  This process is performed because two edges exist in pairs on both sides of the line, and the two edges are recognized as having a reverse sign of the slope and are recognized as a line. This is because the absolute value of the magnitude of the inclination degree of the two edges needs to be close.

That is, the threshold thick_max_x is a threshold related to the thickness of the line.
When the threshold value thick_max_x is set large, the thickness regarded as a line also increases.

  The threshold value thick_max_x means the maximum value of the line width in the X direction to be detected. The threshold thick_max_x may be freely set according to the thickness of the line to be detected.

  In the present embodiment, thick_max_x = 5 and value_max_x = 10 are set.

  Here, when it is desired to extract the line with a limited thickness, “a feature that analyzes the arrangement of the X-direction characteristic differential value data in the X direction and takes the characteristic differential value A (a value other than 0 in the present embodiment). For a feature pixel b that exists in a range between the threshold thick_min_x and the thickness_max_x pixels in the + X direction or the −X direction of the pixel a, the feature differential value B is opposite to the feature differential value A and positive and negative, and an absolute value | A | When the difference from | B | is within the threshold value_max_x, the characteristic pixels a and b are determined to be line edges, and the characteristic pixels a and b are set as values indicating the line information as Y-direction line data. You can do that.

  Of course, “by analyzing the arrangement of the X direction characteristic differential value data in the X direction, the characteristic pixel A taking a characteristic differential value A (a value other than 0 in this embodiment) is greater than or equal to the threshold thick_min_x pixel in the + X direction or the −X direction of the characteristic pixel a. If the characteristic differential value B is opposite to the characteristic differential value A, and the difference between the absolute value | A | and the absolute value | B | is within the threshold value_max_x, the characteristic pixel a And b are determined as line edges, and the characteristic pixels a and b are set as values indicating the line information as Y-direction line data. " It can be set.

  The Y direction line data buffer 210 stores the Y direction line data given from the Y direction line extraction processing unit 209.

  The line extraction processing device 202 outputs the Y direction line data stored in the Y direction line data buffer 210 to the line data output device 203.

FIG. 7 is a flowchart showing the flow of processing performed by the line extraction processing device 202.
With reference to FIG. 7, the flow of processing performed by the above-described line extraction processing device 202 will be described. The processing shown in the flowchart of FIG. 7 is realized by the CPU 101 functioning as a control unit (not shown in FIG. 2) reading out and executing a program stored in the storage device, and controlling each unit shown in FIG. Is done. Note that the process shown in the flowchart of FIG. 7 is merely a specific example of the line extraction process, and may be executed by changing the order of the steps, for example.

  In step S <b> 302, the input image data buffer 204 stores the image data input from the image input device 201 in accordance with a processing start instruction from the control unit.

  Next, in step S <b> 303, the X-direction differentiation processing unit 205 reads the image data input from the input image data buffer 204. Then, differential processing is performed in the X direction, and X direction differential value data is given to the X direction differential value data buffer 206.

  Subsequently, in step S304, the X-direction differential value data buffer 206 stores the X-direction differential value data provided from the X-direction differential processing unit 205.

  In step S 305, the X-direction feature differential value data generation processing unit 207 reads X-direction differential value data from the X-direction differential value data buffer 206. Then, X direction characteristic differential value data is created, and X direction characteristic differential value data is given to the X direction characteristic differential value data buffer 208.

  Next, in step S306, the X-direction feature differential value data buffer 208 stores the X-direction feature differential value data supplied from the X-direction feature differential value data generation processing unit 207.

  Subsequently, in step S307, the Y-direction line extraction processing unit 209 reads X-direction feature differential value data from the X-direction feature differential value data buffer 208. Then, Y direction line data is created, and Y direction line data is given to the Y direction line data buffer 210.

  Finally, in step S308, the Y-direction line data buffer 210 stores the Y-direction line data given from the Y-direction line extraction processing unit 209.

  As described above, the line extraction processing device 202 analyzes the arrangement of pixel values in the X direction, and extracts a line component in the Y direction orthogonal to the X direction.

FIG. 8 is a diagram illustrating a specific example of processing performed by the line extraction processing device 202.
With reference to FIG. 8, processing performed by the line extraction processing device 202 will be described using an example.

  First, the line extraction processing device 202 performs X-direction differential processing and X-direction characteristic differential value data generation processing on image data as shown in P701 to obtain X-direction characteristic differential value data (steps S303 to S303 in FIG. 7). Equivalent to 306). P702 is a diagram showing, based on X-direction feature differential value data, feature pixels that take positive feature differential values with solid lines and feature pixels that take negative feature differential values with broken lines.

  Then, the arrangement in the X direction of the X direction characteristic differential value data is analyzed, and the characteristic differential value B is the characteristic differential value for the characteristic pixel b existing within a predetermined range of the characteristic pixel a taking the characteristic differential value A other than 0. When A is positive and negative and the difference between the absolute value | A | and the absolute value | B | is within a predetermined value, the feature pixels a and b are determined as line edges (in step S307 in FIG. 7). Equivalent). Note that P703 indicates a feature pixel determined to be an edge of a line. As shown in this figure, it can be seen that the edge of the line is paired with the positive and negative data of the characteristic differential value.

  Then, the Y direction line data shown in P704 is stored in the Y direction line data buffer 210 (corresponding to step S308 in FIG. 7). When it is desired to extract the entire line instead of the line edge, the pixels between the feature pixels that form a pair of positive and negative feature differential values may be set as a line region. If the line area is set in this way, Y-direction line data shown in P705 is obtained.

  As described above, the line extraction processing device according to the present embodiment analyzes the arrangement of pixel values in a certain direction, and separates differential value data into a characteristic differential value and a differential value other than the characteristic differential value. Then, a line component in a direction orthogonal to a certain direction is extracted based on the characteristic differential value data. In one edge represented by many different differential values, the differential value having the maximum absolute value determines the strength of the apparent edge, so that the photographic image data is obtained by performing the processing according to this embodiment. For example, the accuracy of the line extraction processing can be improved and the processing speed can be increased.

  Moreover, the line extraction processing apparatus according to the present embodiment can stably detect a line from a strong edge to a weak line by taking into account the magnitude of the absolute value of the differential value.

[Embodiment 2]
The line extraction processing apparatus according to Embodiment 2 extracts line data in two directions, and creates line data by synthesizing the line extraction results in the two directions.

  FIG. 9 is a block diagram illustrating a specific example of a functional configuration for functioning as the line extraction processing device 802 in the computer 1 according to the present embodiment and performing line extraction processing.

  With reference to FIG. 9, the functional configuration of the line extraction processing apparatus 802 according to the second embodiment will be described.

  In addition to the configuration of the line extraction processing device 202, the above function of the line extraction processing device 802 includes a Y-direction differential processing unit 811, a Y-direction differential value data buffer 812, a Y-direction feature differential value data generation processing unit 813, A direction feature differential value data buffer 814, an X direction line extraction processing unit 815, an X direction line data buffer 816, each direction line data composition processing unit 817, and a line data buffer 818 are configured. Other configurations are the same as those of the line extraction processing device 202.

  These components can be configured by a dedicated LSI and a memory that mainly realize the function of the line extraction processing device. However, it may be configured by the CPU 101 of the computer 1 or other hardware. Note that such a dedicated LSI functions as a Y-direction differentiation processing unit 811, a Y-direction feature differential value data generation processing unit 813, and an X-direction line extraction processing unit 815.

  The Y direction differential value data buffer 812, the Y direction characteristic differential value data buffer 814, and the X direction line data buffer 816 are realized by the ROM 103, the RAM 105, the HDD 107, or the like.

  Below, the said component of the line extraction processing apparatus 802 is demonstrated. Note that description of the same components as those of the line extraction processing device 202 is omitted.

  The Y-direction differential processing unit 811 reads image data from the input image data buffer 204. Then, differential processing is performed in the Y direction, and the differential value data in the Y direction is supplied to the Y direction differential value data buffer 812. Note that the Y-direction differentiation processing performed by the Y-direction differentiation processing unit 811 may be any process as long as the degree of the inclination of the pixel value in the Y direction and the rate of change can be understood. In the present embodiment, for example, a filter shown in FIG.

  Returning to FIG. 9, the Y-direction differential value data buffer 812 stores the Y-direction differential value data provided from the Y-direction differential processing unit 811.

  The Y direction characteristic differential value data generation processing unit 813 reads Y direction differential value data from the Y direction differential value data buffer 812. Then, Y-direction feature differential value data created by extracting the Y-direction feature differential value is provided to the Y-direction feature differential value data buffer 814. The Y-direction characteristic differential value data can also be created in the same manner as the X-direction characteristic differential value data.

  The Y-direction feature differential value data buffer 814 stores the Y-direction feature differential value data given from the Y-direction feature differential value data generation processing unit 813.

  The X direction line extraction processing unit 815 reads the Y direction characteristic differential value data from the Y direction characteristic differential value data buffer 814 and supplies the X direction line data to the X direction line data buffer 816.

Here, the X direction line data will be described.
The X-direction line extraction processing unit 815 analyzes the arrangement in the Y direction of the Y-direction feature differential value data. For the feature pixel b existing within the threshold len_th_y pixel in the + Y direction or the −Y direction of the feature pixel a having a feature derivative value A other than 0, the feature differential value B of the feature pixel b is positive or negative. Conversely, if the difference between the absolute value | A | of the feature differential value A and the absolute value | B | of the feature differential value B is within the threshold value_th_y, the feature pixels a and b are determined as line edges, and X The direction line data is set to a value (for example, 1) indicating that the feature pixels a and b are line information. A value indicating line information can be arbitrarily set.

  The X-direction line data buffer 816 stores the X-direction line data given from the X-direction line extraction processing unit 815.

  Each direction line data composition processing unit 817 reads Y direction line data from the Y direction line data buffer 210 and X direction line data from the X direction line data buffer 816. Then, at least one of the X-direction line data and the Y-direction line data is a line area, and the line data set as a line area is supplied to the line data buffer 818.

  The line data buffer 818 stores the line data given from each direction line data composition processing unit 817.

  The line extraction processing device 802 outputs the line data stored in the line data buffer 818 to the line data output device 203.

FIG. 10 is a flowchart showing the flow of processing performed by the line extraction processing device 802.
With reference to FIG. 10, the flow of processing performed by the above-described line extraction processing device 802 will be described. The processing shown in the flowchart of FIG. 10 is realized by the CPU 101 functioning as a control unit (not shown in FIG. 9) reading out and executing a program stored in the storage device and controlling each unit shown in FIG. Is done. Note that the process shown in the flowchart of FIG. 10 is merely a specific example of the line extraction process, and may be executed by changing the order of the steps, for example.

  Moreover, since the process of step S902-908 is the same as the process of step S302-308 of FIG. 7, description is abbreviate | omitted.

  In step S909, the Y-direction differentiation processing unit 811 reads image data input from the input image data buffer 204. Then, differential processing is performed in the Y direction, and Y direction differential value data is given to the Y direction differential value data buffer 812.

  Subsequently, in step S910, the Y-direction differential value data buffer 812 stores the Y-direction differential value data provided from the Y-direction differential processing unit 811.

  In step S911, the Y-direction characteristic differential value data generation processing unit 813 reads the Y-direction differential value data from the Y-direction differential value data buffer 812. Then, Y-direction feature differential value data is created, and Y-direction feature differential value data is given to the Y-direction feature differential value data buffer 814.

  Next, in step S912, the Y-direction feature differential value data buffer 814 stores the Y-direction feature differential value data provided from the Y-direction feature differential value data generation processing unit 813.

  Subsequently, in step S <b> 913, the X-direction line extraction processing unit 815 reads Y-direction feature differential value data from the Y-direction feature differential value data buffer 814. Then, X direction line data is created, and X direction line data is given to the X direction line data buffer 816.

  In step S914, the X-direction line data buffer 816 stores the X-direction line data given from the X-direction line extraction processing unit 815.

  Next, in step S <b> 915, each direction line data composition processing unit 817 reads Y direction line data from the Y direction line data buffer 210 and X direction line data from the X direction line data buffer 816. Then, at least one of the X-direction line data and the Y-direction line data is a line area, and the line data set as a line area is supplied to the line data buffer 818.

  Finally, in step S916, the line data buffer 818 stores the line data provided from each direction line data composition processing unit 817.

  As described above, the line extraction processing device 802 extracts lines in two directions, and synthesizes the line extraction results in the two directions.

FIG. 11 is a diagram illustrating a specific example of processing performed by the line extraction processing device 802.
Referring to FIG. 11, the process performed by line extraction processing device 802 in steps S909 to 914 in FIG. 10 will be described using an example.

  The line extraction processing device 802 obtains Y-direction characteristic differential value data by performing Y-direction differential processing and Y-direction characteristic differential value data generation processing on the image data as shown in P701 (corresponding to steps S909 to 912 in FIG. 10). ). P1002 is a diagram showing, based on Y-direction feature differential value data, feature pixels that take positive feature differential values with solid lines, and feature pixels that take negative feature differential values with broken lines.

  Then, the Y-direction characteristic differential value data is analyzed in the Y direction, and the characteristic differential value B is the characteristic differential value for the characteristic pixel b existing within the predetermined range of the characteristic pixel a taking the characteristic differential value A other than 0. If A is positive and negative and the difference between the absolute value | A | and the absolute value | B | is within a predetermined value, the feature pixels a and b are determined to be line edges (in step S913 in FIG. 10). Equivalent). Note that P1003 indicates a feature pixel determined to be an edge of a line. As shown in this figure, it can be seen that the edge of the line is paired with the positive and negative data of the characteristic differential value.

  Then, the X direction line data indicated by P1004 is stored in the X direction line data buffer 816 (corresponding to step S914 in FIG. 10). When it is desired to extract the entire line instead of the line edge, the pixels between the feature pixels that form a pair of positive and negative feature differential values may be set as a line region. If the line area is set in this way, the X-direction line data shown in P1005 is obtained.

  FIG. 12 is a diagram illustrating a specific example of processing (extracting line edges) performed by the line extraction processing device 802.

  With reference to FIG. 12, the process of extracting line edges by the line extraction processing device 802 in steps S915 and 916 of FIG. 10 will be described using an example.

  Each direction line data composition processing unit 817 reads Y direction line data indicated by P704 from the Y direction line data buffer 210 and X direction line data indicated by P1004 from the X direction line data buffer 816. Then, the line data shown in P1103 is created in which at least one of the X-direction line data and the Y-direction line data is a line area (corresponding to step S915 in FIG. 10), and the line data buffer 818 is created. (Corresponding to step S916 in FIG. 10).

  FIG. 13 is a diagram illustrating a specific example of the process (extracting the entire line) performed by the line extraction processing device 802.

  In FIG. 13, as described above, when it is desired to extract the entire line, the Y direction line data indicated by P705 and the X direction indicated by P1005 are set by setting a line area between the positive / negative data of the characteristic differential value. Line data is obtained.

  Each direction line data composition processing unit 817 creates line data shown in P1203 in which at least one of the X direction line data and the Y direction line data is a line area (step S915 in FIG. 10). Stored in the line data buffer 818 (corresponding to step S916 in FIG. 10).

  As described above, the line extraction processing device according to the present embodiment analyzes the arrangement of pixel values in a certain direction, and separates differential value data into a characteristic differential value and a differential value other than the characteristic differential value. Then, a line component in a direction orthogonal to a certain direction is extracted based on the characteristic differential value data. Such processing is performed in two directions, and line data is created by combining the line extraction results in the two directions. Thereby, in addition to the effects of the first embodiment, the accuracy of line extraction processing in two directions can be improved and the processing speed can be increased.

[Embodiment 3]
In addition to the processing of the line extraction processing device according to the first embodiment, the line extraction processing device according to the third embodiment adds an analysis in the direction perpendicular to the direction in which the characteristic differential value data is differentiated. To prevent.

  FIG. 14 is a block diagram illustrating a specific example of a functional configuration that functions as the line extraction processing device 1302 in the computer 1 according to the present embodiment and performs line extraction processing.

  With reference to FIG. 14, the functional configuration of the line extraction processing apparatus 1302 according to the second embodiment will be described.

  The above-described functions of the line extraction processing device 1302 include an X-direction feature differential value data first buffer 1308 and an X-direction feature differential value data noise removal processing unit instead of the X-direction feature differential value data buffer 208 of the line extraction processing device 202. 1309 and an X-direction feature differential value data second buffer 1310. Other configurations are the same as those of the line extraction processing device 202.

  These components can be configured by a dedicated LSI and a memory that mainly realize the function of the line extraction processing device. However, it may be configured by the CPU 101 of the computer 1 or other hardware. Such a dedicated LSI functions as the X-direction feature differential value data noise removal processing unit 1309.

  The X-direction feature differential value first data buffer 1308 and the X-direction feature differential value data second buffer 1310 are realized by the ROM 103, the RAM 105, the HDD 107, or the like.

  The above components of the line extraction processing device 1302 will be described below. Note that description of the same components as those of the line extraction processing device 202 is omitted.

  The X-direction feature differential value data first buffer 1308 stores the X-direction feature differential value data provided from the X-direction feature differential value data generation processing unit 207.

  The X-direction feature differential value data noise removal processing unit 1309 reads X-direction feature differential value data from the X-direction feature differential value data first buffer 1308. Then, unnecessary noise in the X-direction feature differential value data is removed, and the X-direction feature differential value data is supplied to the X-direction feature differential value data second buffer 1310.

Here, the noise removal processing of the X direction characteristic differential value will be described.
The noise removal processing performed by the X-direction feature differential value data noise removal processing unit 1309 is X-direction feature differential value data stored in the X-direction feature differential value data first buffer 1308, and is a short line that is desired to be removed as a line. The detection accuracy is improved by removing the X-direction characteristic differential value of the point and the point from the characteristic differential value (in this embodiment, zero).

  Specifically, the data string in the Y direction of the X direction characteristic differential value is analyzed. When the absolute value of the difference between the characteristic differential values of the feature pixels adjacent in the Y direction is equal to or smaller than the threshold value CONT_VAL_TH_X, the number of connections is counted on the basis that the feature pixels adjacent in the Y direction are connected as a line. Then, the connected X-direction feature differential values whose connection number is equal to or less than the threshold value CONT_TH_X are excluded from the feature differential values.

  The threshold value CONT_VAL_TH_X is a value for determining whether or not feature differential values are connected, and an arbitrary positive real number can be set as the value. In this embodiment, CONT_VAL_TH_X = 10.

  CONT_TH_X indicates the minimum size of the line thickness extracted in the Y direction. Therefore, it may be set appropriately according to the purpose of detection, and may be any positive real number. In the present embodiment, CONT_VAL_TH_X = 5.

Thereby, since the thing with a short line and a point are not extracted as a line, the precision of line extraction improves.
The X-direction feature differential value data second buffer 1310 stores the X-direction feature differential value data that has been subjected to noise removal, provided from the X-direction feature differential value data noise removal generation processing unit 1309.

  The Y direction line extraction processing unit 209 reads X direction characteristic differential value data from the X direction characteristic differential value data second buffer 1310. Then, the Y direction line data is supplied to the Y direction line data buffer 210.

  FIG. 15 is a flowchart showing the flow of processing performed by the line extraction processing device 1302.

  With reference to FIG. 15, the flow of processing performed by the above-described line extraction processing device 1302 will be described. The process shown in the flowchart of FIG. 15 is realized by the CPU 101 functioning as a control unit (not shown in FIG. 14) reading out and executing a program stored in the storage device and controlling each unit shown in FIG. Is done. The process shown in the flowchart of FIG. 15 is only a specific example of the line extraction process, and may be executed by changing the order of each step, for example.

  Further, the processing of steps S1402 to 1405 is the same as the processing of steps S302 to 305 in FIG.

  In step S1406, the X-direction feature differential value data first buffer 1308 stores the X-direction feature differential value data provided from the X-direction feature differential value data generation processing unit 207.

  In step S 1407, the X-direction feature differential value data noise removal processing unit 1309 reads X-direction feature differential value data from the X-direction feature differential value data first buffer 1308. Then, unnecessary noise in the X-direction feature differential value data is removed, and the X-direction feature differential value data is supplied to the X-direction feature differential value data second buffer 1310.

  Subsequently, in step S1408, the X-direction feature differential value data second buffer 1310 stores the X-direction feature differential value data from which noise has been removed.

  In step S1409, the Y-direction line extraction processing unit 209 reads X-direction feature differential value data from the X-direction feature differential value data second buffer 1310. Then, Y direction line data is created and given to the Y direction line data buffer 210.

  Finally, in step S1410, the Y-direction line data buffer 210 stores the line data provided from the Y-direction line extraction processing unit 209.

  As described above, the line extraction processing apparatus according to this embodiment adds analysis in the direction perpendicular to the differentiated direction of the characteristic differential value data. Thereby, it is possible to perform detection in a form in which a very short line that should not be recognized as a dot or line of halftone dots is removed from the line extraction processing apparatus described in the first embodiment. Therefore, detection accuracy can be improved and line extraction errors can be prevented.

[Embodiment 4]
In addition to the processing of the line extraction processing device according to the second embodiment, the line extraction processing device according to the fourth embodiment performs noise removal processing as described in the third embodiment in each of the two directions to extract lines, respectively. Line data is created by combining the line extraction results in the two directions.

  FIGS. 16 and 17 are block diagrams showing specific examples of a functional configuration for functioning as the line extraction processing device 1502 in the computer 1 according to the present embodiment and performing the line extraction processing. Note that FIG. 16 is also a part of a functional configuration of a line extraction processing device 2202 described later.

  With reference to FIGS. 16 and 17, the functional configuration of the line extraction processing apparatus 1502 according to Embodiment 4 will be described.

  The function of the line extraction processing device 1502 is to replace the X-direction feature differential value data buffer 208 and the Y-direction feature differential value data buffer 814 of the line extraction processing device 802 with an X-direction feature differential value data first buffer 1308, X Direction characteristic differential value data noise elimination processing unit 1309, X direction characteristic differential value data second buffer 1310, Y direction characteristic differential value data first buffer 1513, Y direction characteristic differential value data noise elimination processing unit 1514, Y A direction characteristic differential value data second buffer 1515 is provided. Other configurations are the same as those of the line extraction processing device 802. The X-direction feature differential value data first buffer 1308, the X-direction feature differential value data noise removal processing unit 1309, and the X-direction feature differential value data second buffer 1310 are configured in the line extraction processing device 1302 according to the third embodiment. Same as element.

  These components can be configured by a dedicated LSI and a memory that mainly realize the function of the line extraction processing device. However, it may be configured by the CPU 101 of the computer 1 or other hardware. Such a dedicated LSI functions as the Y-direction feature differential value data noise removal processing unit 1514.

  The Y-direction characteristic differential value first data buffer 1513 and the Y-direction characteristic differential value data second buffer 1515 are realized by the ROM 103, the RAM 105, the HDD 107, or the like.

  The above components of the line extraction processing device 1502 will be described below. Note that description of the same components as those of the line extraction processing devices 802 and 1302 is omitted.

  The Y-direction feature differential value data first buffer 1513 stores the Y-direction feature differential value data given from the Y-direction feature differential value data generation processing unit 813.

  The Y-direction feature differential value data noise removal processing unit 1514 reads Y-direction feature differential value data from the Y-direction feature differential value data first buffer 1512. Then, unnecessary noise in the Y-direction characteristic differential value data is removed, and the Y-direction characteristic differential value data is supplied to the Y-direction characteristic differential value data second buffer 1515.

Here, the noise removal processing of the Y-direction feature differential value will be described.
The noise removal processing performed by the Y-direction characteristic differential value data noise removal processing unit 1514 is Y-direction characteristic differential value data stored in the Y-direction characteristic differential value data first buffer 1513, and is a short line to be removed as a line. The detection accuracy is increased by removing the Y-direction characteristic differential value of the point and the point from the characteristic differential value (in this embodiment, zero).

  Specifically, the data string in the X direction of the Y direction characteristic differential value is analyzed. When the absolute value of the difference between the feature differential values of features adjacent in the X direction is equal to or less than the threshold value CONT_VAL_TH_Y, the number of connections is counted on the basis that the feature differential values adjacent in the X direction are connected. Then, the connected Y-direction feature differential value whose connection number is equal to or less than the threshold value CONT_TH_Y is excluded from the feature differential value.

  The threshold value CONT_VAL_TH_Y is a value for determining whether or not feature differential values are connected, and an arbitrary positive real number can be set as the value. In this embodiment, CONT_VAL_TH_Y = 10.

  CONT_TH_Y indicates the minimum size of the line thickness extracted in the X direction. Therefore, it may be set appropriately according to the purpose of detection, and may be any positive real number. In the present embodiment, CONT_VAL_TH_Y = 5.

Thereby, since the thing with a short line and a point are not extracted as a line, the precision of line extraction improves.
The Y-direction feature differential value data second buffer 1515 stores Y-direction feature differential value data from the Y-direction feature differential value data noise removal generation processing unit 1514 that has been subjected to noise removal.

  The X-direction line extraction processing unit 815 reads Y-direction feature differential value data from the Y-direction feature differential value data second buffer 1515. Then, the X direction line data is supplied to the X direction line data buffer 816.

  FIG. 18 is a flowchart showing the flow of processing performed by the line extraction processing device 1502.

  With reference to FIG. 18, the flow of processing performed by the above-described line extraction processing device 1502 will be described. In the process shown in the flowchart of FIG. 18, the CPU 101 functioning as a control unit (not shown in FIGS. 16 and 17) reads and executes a program stored in the storage device, and controls each unit shown in FIGS. It is realized by doing. Note that the process shown in the flowchart of FIG. 18 is merely a specific example of the line extraction process, and may be executed by changing the order of the steps, for example.

  Further, the processing in steps S1702 to 1710 is the same as the processing in steps S1402 to 1410 in FIG. 15, and the processing in steps S1711 to 1713 and 1717 to 1720 is the same as the processing in steps S909 to 916 in FIG. Omitted.

  In step S1714, the Y-direction feature differential value data first buffer 1513 stores the Y-direction feature differential value data provided from the Y-direction feature differential value data generation processing unit 813.

  Next, in step S1715, the Y-direction feature differential value data noise removal processing unit 1513 reads Y-direction feature differential value data from the Y-direction feature differential value data first buffer 1513. Then, unnecessary noise in the Y-direction characteristic differential value data is removed, and the Y-direction characteristic differential value data is supplied to the Y-direction characteristic differential value data second buffer 1515.

  Subsequently, in step S1716, the Y-direction feature differential value data second buffer 1515 stores the Y-direction feature differential value data from which noise has been removed.

  Since the processing after step S1717 is the same as that of the line extraction processing device 802 as described above, the description thereof is omitted.

  As described above, the line extraction processing apparatus according to the present embodiment performs noise removal processing in two directions as described in the third embodiment to extract lines, and synthesizes the line extraction results in the two directions. Thus, line data is created. As a result, the line extraction processing apparatus described in the second embodiment does not extract a short line or a point as a line, thereby improving the line extraction accuracy.

[Embodiment 5]
The line extraction processing apparatus according to the fifth embodiment performs line extraction using a combination of X direction characteristic differential value data and Y direction characteristic differential value data. Thereby, compared with the line extraction processing apparatus which concerns on Embodiment 2, the detection accuracy of the line of directions other than a X direction and a Y direction can be improved.

  19 and 20 are block diagrams showing specific examples of a functional configuration for functioning as the line extraction processing device 1902 in the computer 1 according to the present embodiment and performing line extraction processing.

  With reference to FIGS. 19 and 20, the functional configuration of the line extraction processing apparatus 1902 according to Embodiment 5 will be described.

  In addition to the configuration of the line extraction processing device 802, the above function of the line extraction processing device 1902 includes each direction feature differential value data synthesis processing unit 1913 and a feature differential value data buffer 1914. Other configurations are the same as those of the line extraction processing device 802.

  These components can be configured by a dedicated LSI and a memory that mainly realize the function of the line extraction processing device. However, it may be configured by the CPU 101 of the computer 1 or other hardware. Such a dedicated LSI functions as each direction feature differential value data synthesis processing unit 1913.

  The feature differential value data buffer 1914 is realized by the ROM 103, the RAM 105, the HDD 107, or the like.

  The above components of the line extraction processing device 1902 will be described below. Note that description of the same components as those of the line extraction processing device 802 is omitted.

  Each direction characteristic differential value data composition processing unit 1913 reads X direction characteristic differential value data from the X direction characteristic differential value data buffer 208 and Y direction characteristic differential value data from the Y direction characteristic differential value data buffer 814, respectively. Then, the combined characteristic differential value data is given to the characteristic differential value data buffer 1914.

  Here, the “composite feature differential value data” given by each direction feature differential value data 1913 is obtained by adding the feature differential values in each direction of the X direction feature differential value data and the Y direction feature differential value data.

  In the line extraction processing apparatus according to Embodiment 2, the process of extracting lines is performed independently in each direction. For this reason, a line in a direction different from each direction (for example, a line in an oblique direction when each direction is a vertical or horizontal direction) is difficult to be detected because values are dispersed in characteristic differential values in each direction. There was a trend.

  However, in the line extraction processing apparatus of the present embodiment, lines other than the X direction and the Y direction can be detected more accurately by adding the characteristic differential values in each direction.

  In addition, when intentionally placing importance on the X-direction and Y-direction lines, the characteristic differential value data may be a value having a large characteristic differential value in each direction of the X-direction characteristic differential value data and the Y-direction characteristic differential value data. Good.

  The composite feature differential value data buffer 1914 stores the composite feature differential value data given from each direction feature differential value data composition processing unit 1913.

  FIG. 21 is a flowchart showing the flow of processing performed by the line extraction processing device 1902.

  With reference to FIG. 21, the flow of processing performed by the above-described line extraction processing device 1902 will be described. In the processing shown in the flowchart of FIG. 21, the CPU 101 functioning as a control unit (not shown in FIGS. 19 and 20) reads out and executes a program stored in the storage device, and controls each unit shown in FIGS. It is realized by doing. The process shown in the flowchart of FIG. 21 is only a specific example of the line extraction process, and may be executed by changing the order of the steps, for example.

  Moreover, since the process of step S2102-2106, 2107-2110 is the same as the process of step S902-906 of FIG. 10, and 909-912, description is abbreviate | omitted.

  In step S <b> 2111, each directional feature differential value data composition processing unit 1913 reads X-direction feature differential value data from the X-direction feature differential value data buffer 1908 and Y-direction feature differential value data buffer 1912. . Then, the combined feature differential value data obtained by adding the feature differential value data in each direction is supplied to the combined feature differential value data buffer 1914.

  Next, in step S2112, the synthesized feature differential value data buffer 1914 stores the synthesized feature differential value data provided from each direction feature differential value data synthesis processing unit 1913.

  In the processing after step S2113, the Y-direction line extraction processing unit 209 and the X-direction line extraction processing unit 815 read the combined feature differential value data from the combined feature differential value data buffer 1914, and convert the Y-direction and X-direction line data. Except for giving to each buffer, the processing is the same as the processing in steps S907 to 908 and 913 to 916 in FIG.

  As described above, the line extraction processing apparatus according to the present embodiment performs line extraction using a combination of X-direction feature differential value data and Y-direction feature differential value data. Thereby, the detection accuracy of lines in directions other than the X direction and the Y direction can be improved.

[Embodiment 6]
The line extraction processing apparatus according to the sixth embodiment performs line extraction using the synthesized noise-removed X-direction characteristic differential value data and Y-direction characteristic differential value data. Thereby, the detection accuracy of lines in directions other than the X direction and the Y direction can be improved as compared with the line extraction processing apparatus according to the fourth embodiment.

  FIG. 22 is a block diagram specifically showing a part of a functional configuration for functioning as the line extraction processing device 2202 in the computer 1 according to the present embodiment and performing the line extraction processing.

  With reference to FIGS. 16 and 22, the functional configuration of the line extraction processing apparatus 2202 according to Embodiment 6 will be described.

  The above-described function of the line extraction processing device 2202 includes the components shown in FIG. 16 and the components shown in FIG.

  Each directional feature differential value data synthesis processing unit 1913 removes noise from the X-direction feature differential value data second buffer 1310 and removes noise from the Y-direction feature differential value data second buffer 1515. The Y direction characteristic differential value data is read. Then, the composite feature differential value data 1914 is given to the composite feature differential value data buffer 1914.

  As described in the fifth embodiment, the combined feature differential value data is obtained by adding the directional feature differential values of the X-direction feature differential value data and the Y-direction feature differential value data.

  In the line extraction processing apparatus according to the fourth embodiment, the process of extracting the noise-removed line is performed independently in each direction. For this reason, a line in a direction different from each direction (for example, a line in an oblique direction when each direction is a vertical or horizontal direction) is difficult to be detected because values are dispersed in characteristic differential values in each direction. There was a trend.

  However, in the line extraction processing apparatus of the present embodiment, lines other than the X direction and the Y direction can be detected more accurately by adding the characteristic differential values in each direction.

  In addition, when intentionally placing importance on the X-direction and Y-direction lines, the characteristic differential value data may be a value having a large characteristic differential value in each direction of the X-direction characteristic differential value data and the Y-direction characteristic differential value data. Good.

  23 and 24 are flowcharts showing the flow of processing performed by the line extraction processing device 2202.

  With reference to FIGS. 23 and 24, the flow of processing performed by the above-described line extraction processing device 2202 will be described. In the process shown in the flowchart of FIG. 22, CPU 101 functioning as a control unit (not shown in FIGS. 16 and 22) reads and executes a program stored in the storage device, and controls each unit shown in FIGS. 16 and 22. It is realized by doing. The process shown in the flowchart of FIG. 23 is merely a specific example of the line extraction process, and may be executed by changing the order of the steps, for example.

  Moreover, since the process of step S2402-2408 and 2409-2414 is the same as the process of step S1702-1708 and 1711-1716 of FIG. 18, description is abbreviate | omitted.

  In step S 2415, each directional feature differential value data composition processing unit 1913 receives the X-direction feature differential value data from which noise has been removed from the X-direction feature differential value data second buffer 1310 from the Y-direction feature differential value data second buffer 1515. Read Y-direction feature differential value data from which noise has been removed. Then, the combined feature differential value data obtained by adding the feature differential value data in each direction is supplied to the combined feature differential value data buffer 1914.

  Since the process after step S2416 is the same as the process of step S212-1218 of FIG. 21, description is abbreviate | omitted.

  As described above, the line extraction processing apparatus according to the present embodiment performs line extraction using a combination of noise-removed X-direction feature differential value data and Y-direction feature differential value data. Thereby, the detection accuracy of lines in directions other than the X direction and the Y direction can be improved as compared with the line extraction processing apparatus described in the fifth embodiment.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a block diagram which shows the specific example of the hardware constitutions of the computer 1 which mounts the line extraction processing apparatus 202 which concerns on this Embodiment. It is a block diagram which functions as the line extraction processing apparatus 202 in the computer 1 which concerns on this Embodiment, and shows the specific example of a functional structure for performing a line extraction process. It is a figure which shows the filter for calculating the difference with an adjacent pixel. It is a figure for demonstrating a X direction characteristic differential value. It is a figure which shows the characteristic differential value data produced based on the differential value data shown by FIG.4 (c). It is a figure which shows another example of differential value data and characteristic differential value data. It is the flowchart which showed the flow of the process which the line extraction processing apparatus 202 performs. It is a figure which shows the specific example of the process which the line extraction processing apparatus 202 performs. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 802 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. It is the flowchart which showed the flow of the process which the line extraction processing apparatus 802 performs. It is a figure which shows the specific example of the process which the line extraction processing apparatus 802 performs. It is a figure which shows the specific example of the process (The edge of a line is extracted) which the line extraction processing apparatus 802 performs. It is a figure which shows the specific example of the process (The whole line is extracted) which the line extraction processing apparatus 802 performs. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 1302 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. 10 is a flowchart showing a flow of processing performed by the line extraction processing device 1302. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 1502 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 1502 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. 6 is a flowchart showing a flow of processing performed by the line extraction processing device 1502. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 1902 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. It is a block diagram which shows the specific example of a function structure for functioning as the line extraction processing apparatus 1902 in the computer 1 which concerns on this Embodiment, and performing a line extraction process. 5 is a flowchart showing a flow of processing performed by a line extraction processing device 1902; It is a block diagram which specifically shows a part of functional structure for functioning as the line extraction processing device 2202 in the computer 1 according to the present embodiment and performing line extraction processing. 12 is a flowchart showing a flow of processing performed by the line extraction processing device 2202. 12 is a flowchart showing a flow of processing performed by the line extraction processing device 2202.

Explanation of symbols

  1 computer, 2 recording medium, 101 CPU, 103 ROM, 105 RAM, 107 HDD, 109 reading unit, 111 input unit, 113 communication unit, 115 display unit, 201 image input device, 202 line extraction processing device, 203 line data output Device, 204 input image data buffer, 205 X direction differential processing unit, 206 X direction differential value data buffer, 207 X direction characteristic differential value data generation processing unit, 208 X direction characteristic differential value data buffer, 209 Y direction line extraction processing unit 210 Y-direction line data buffer, 802 Line extraction processing device, 811 Y-direction differential processing unit, 812 Y-direction differential value data buffer, 813 Y-direction characteristic differential value data generation processing unit, 814 Y-direction characteristic differential value data buffer, 815 X-direction line extraction processing unit, 816 X-direction line data Buffer, 817 Each direction line data composition processing unit, 818 line data buffer, 1302 line extraction processing device, 1308 X direction feature differential value data first buffer, 1309 X direction feature differential value data noise removal processing unit, 1310 X direction feature differentiation Value data second buffer, 1902 line extraction processing device, 1913 each direction feature differential value data synthesis processing unit, 1914 synthesis feature differential value data buffer, 2202 line extraction processing device.

Claims (13)

A line extraction processing device that extracts data corresponding to a line included in image data from image data corresponding to a coordinate plane formed by a coordinate axis extending in the first direction and a coordinate axis extending in a second direction orthogonal to the axis Because
Image input means for inputting the image data;
First direction differentiation processing means for differentiating the image data in a first direction to generate first direction differential value data;
Generating first direction feature differential value data in which the first direction differential value having the maximum absolute value in a positive or negative continuous section of the first direction differential value data is used as the first direction feature differential value. First direction feature differential value data generation processing means for
And a first direction line extraction processing means for outputting the second direction line data representing information of said second direction to the corresponding line,
The feature pixel taking the first direction differential value includes a first feature pixel whose first direction feature differential value is other than 0,
The first direction line extraction processing means is a second feature that is the feature pixel that exists within a first predetermined range related to the thickness of the line from the position of the first feature pixel in the first direction. The first direction feature differential value of the pixel is opposite in polarity to the first direction feature differential value of the first feature pixel, and the absolute value of the first direction feature differential value of the first feature pixel And the absolute value of the first direction feature differential value of the second feature pixel is a first predetermined value for determining that both pixels correspond to two edges on both sides of the line. A value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set according to the determination that the value is less than or equal to A line extraction processing apparatus for outputting the second direction line data . The first direction line extraction processing unit, the position of the first feature pixel in the previous SL first direction, the characteristic pixel der Ru second present in the first predetermined range for the thickness of the wire in the first direction, wherein the differential value of the feature pixel in the first direction, wherein the differential value and polarity reverse of the first feature pixel, and, in the first direction, wherein the differential value of the first feature pixel A difference between the absolute value and the absolute value of the first direction feature differential value of the second feature pixel is a first for determining that both pixels correspond to two edges on both sides of the line . The first feature pixel, the second feature pixel, and the first feature pixel and the second feature pixel in the first direction. indicating that a pixel is present, a pixel constituting the line in the second direction between the There outputs the set the second direction line data, the line extracting apparatus according to claim 1, wherein. A first direction feature differential value data noise removal processing means for removing noise from the first direction feature differential value data ;
The first direction feature differential value data noise removal processing means includes a difference between the first feature pixel and the first direction feature differential value equal to or less than a second predetermined value. The number of connected feature pixels for the first feature pixel is calculated using the third feature pixel adjacent in the second direction as the feature pixel connected to the first feature pixel. The first direction feature differential value of the feature pixel connected to the first feature pixel and the first feature pixel based on the first direction feature differential value data Remove the data as the noise,
The second predetermined value refers to a value for determining whether or not feature pixels adjacent in the second direction are connected as a line,
It said third predetermined value, the extracted in the second direction refers to the minimum size of the thickness of the line, the line extracting apparatus according to claim 1 or 2 wherein. A line extraction processing device that extracts data corresponding to a line included in image data from image data corresponding to a coordinate plane formed by a coordinate axis extending in the first direction and a coordinate axis extending in a second direction orthogonal to the axis Because
Image input means for inputting the image data;
First direction differentiation processing means for differentiating the image data in a first direction to generate first direction differential value data;
Generating first direction feature differential value data in which the first direction differential value having the maximum absolute value in a positive or negative continuous section of the first direction differential value data is used as the first direction feature differential value. First direction feature differential value data generation processing means for
First direction line extraction processing means for outputting second direction line data representing information of a line corresponding to the second direction;
Second direction differentiation processing means for differentiating the image data in the second direction to create second direction differential value data;
Generating second direction characteristic differential value data in which the second direction differential value having the maximum absolute value in a positive or negative continuous section of the second direction differential value data is used as the second direction characteristic differential value. Second direction feature differential value data generation processing means for
Second direction line extraction processing means for outputting first direction line data representing information of a line corresponding to the first direction;
Each direction line data synthesizing means for creating line data by synthesizing the data output by the first direction line extraction processing means and the second direction line extraction processing means,
The feature pixel taking the first direction differential value includes a first feature pixel whose first direction feature differential value is other than 0,
The first direction line extraction processing means is a second feature that is the feature pixel that exists within a first predetermined range related to the thickness of the line from the position of the first feature pixel in the first direction. The first direction feature differential value of the pixel is opposite in polarity to the first direction feature differential value of the first feature pixel, and the absolute value of the first direction feature differential value of the first feature pixel And the absolute value of the first direction feature differential value of the second feature pixel is a first predetermined value for determining that both pixels correspond to two edges on both sides of the line. A value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set according to the determination that the value is less than or equal to Output the second direction line data,
The second direction line extraction processing means is a third feature that is the feature pixel that exists within a first predetermined range related to the thickness of the line from the position of the first feature pixel in the second direction. The second direction feature differential value of the pixel is opposite in polarity to the second direction feature differential value of the first feature pixel, and the absolute value of the second direction feature differential value of the first feature pixel And the absolute value of the second direction characteristic differential value of the third feature pixel is a second predetermined value for determining that both pixels correspond to two edges on both sides of the line. A value indicating that the first feature pixel and the third feature pixel are pixels constituting an edge of the line in the first direction. were you output the first direction line data, the line extraction process unit. For synthesizing the first direction characteristic differential value data and the second direction characteristic differential value data to generate combined characteristic differential value data indicating the region of the line in the coordinate plane corresponding to the image data The line extraction processing device according to claim 4 , further comprising a directional feature differential value data synthesis processing unit. The first direction line extraction processing means is a second feature that is the feature pixel that exists within a first predetermined range related to the thickness of the line from the position of the first feature pixel in the first direction. The first direction feature differential value of the pixel is opposite in polarity to the first direction feature differential value of the first feature pixel, and the absolute value of the first direction feature differential value of the first feature pixel And the absolute value of the first direction feature differential value of the second feature pixel is a first predetermined value for determining that both pixels correspond to two edges on both sides of the line. Between the first feature pixel, the second feature pixel, and the first feature pixel and the second feature pixel in the first direction. Indicates that the pixel that constitutes the line in the second direction is a pixel existing in There outputs set the second direction line data,
The second direction line extraction processing means is a third feature that is the feature pixel that exists within a first predetermined range related to the thickness of the line from the position of the first feature pixel in the second direction. The second direction feature differential value of the pixel is opposite in polarity to the second direction feature differential value of the first feature pixel, and the absolute value of the second direction feature differential value of the first feature pixel And the absolute value of the second direction characteristic differential value of the third feature pixel is a second predetermined value for determining that both pixels correspond to two edges on both sides of the line. Between the first feature pixel, the third feature pixel, and the first feature pixel and the third feature pixel in the first direction. Indicates that the pixels in the first direction constitute the line in the first direction. There outputs the set first direction line data, the line extracting apparatus according to claim 4 or 5. First direction feature differential value data noise removal processing means for removing noise from the first direction feature differential value data;
At least one of second direction feature differential value data noise removal processing means for removing noise from the second direction feature differential value data;
The first direction feature differential value data noise removal processing means includes the first feature pixel and the first feature pixel whose difference from the first direction feature differential value of the first feature pixel is a third predetermined value or less. A feature pixel that is adjacent in the second direction is a feature pixel that is connected to the first feature pixel, and the number of connected feature pixels for the first feature pixel is calculated. The first direction of the feature pixel connected in the second direction to the first feature pixel and the first feature pixel from the first direction feature differential value data according to the determination that the value is less than or equal to the value Characteristic differential value data is removed as the noise to obtain the first direction characteristic differential value data,
The second direction feature differential value data noise removal processing means includes the first feature pixel and a difference between the first feature pixel and the second direction feature differential value not more than a fifth predetermined value. A feature pixel that is adjacent in the first direction is a feature pixel that is connected to the first feature pixel, and the number of connected feature pixels for the first feature pixel is calculated, and the number of connections is a sixth predetermined number of pixels. The second direction of the feature pixel connected to the first feature pixel and the first feature pixel in the first direction from the second direction feature differential value data in response to the determination that the value is less than or equal to the value. The characteristic differential value is removed to obtain the second direction characteristic differential value data ,
The third predetermined value refers to a value for determining whether or not feature pixels adjacent in the second direction are connected as a line,
The fourth predetermined value refers to a minimum size of the thickness of the line extracted in the second direction;
The fifth predetermined value refers to a value for determining whether or not feature pixels adjacent in the first direction are connected as a line;
The line extraction processing apparatus according to claim 4 , wherein the sixth predetermined value indicates a minimum size of a thickness of the line extracted in the first direction . A program for causing a computer having a calculation unit to execute processing for extracting data corresponding to a line included in image data,
The image data is image data corresponding to a coordinate plane formed by a coordinate axis extending in a first direction and a coordinate axis extending in a second direction orthogonal to the axis,
The arithmetic unit inputting the image data;
The computing unit differentiating the image data in a first direction to create first direction differential value data;
A first direction feature having the first direction differential value having a maximum absolute value in a section where the differential value is positive or negative in the first direction differential value data as the first direction feature differential value. Generating differential value data; and
The arithmetic unit comprising: outputting second direction line data representing information of a line corresponding to the second direction;
The feature pixel taking the first direction differential value includes a first feature pixel whose first direction feature differential value is other than 0,
In the step of outputting the second direction line data,
The first direction feature differential value of the second feature pixel which is the feature pixel existing within the first predetermined range regarding the thickness of the line from the position of the first feature pixel in the first direction is , The first direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the first direction feature differential value of the first feature pixel and the second feature pixel of the second feature pixel are the same. According to determination that the difference between the absolute value of the one-way characteristic differential value is equal to or less than a first predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the second direction line data in which a value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set. , Line extraction processing program. A program for causing a computer having a calculation unit to execute processing for extracting data corresponding to a line included in image data,
The image data is image data corresponding to a coordinate plane formed by a coordinate axis extending in a first direction and a coordinate axis extending in a second direction orthogonal to the axis,
The arithmetic unit inputting the image data;
The computing unit differentiating the image data in a first direction to create first direction differential value data;
A first direction feature having the first direction differential value having a maximum absolute value in a section where the differential value is positive or negative in the first direction differential value data as the first direction feature differential value. Generating differential value data; and
The operation unit outputting second direction line data representing information of a line corresponding to the second direction;
The computing unit differentiating the image data in the second direction to create second direction differential value data;
In the second direction feature, the computing unit sets the second direction differential value having the maximum absolute value in the positive or negative continuous section of the second direction differential value data as the second direction feature differential value. Generating differential value data; and
The computing unit outputting first direction line data representing information of a line corresponding to the first direction;
The arithmetic unit comprises the step of generating line data by combining the data output by the step of outputting the first direction line data and the step of outputting the second direction line data;
The feature pixel taking the first direction differential value includes a first feature pixel whose first direction feature differential value is other than 0,
In the step of outputting the first direction line data,
The first direction feature differential value of the second feature pixel which is the feature pixel existing within the first predetermined range regarding the thickness of the line from the position of the first feature pixel in the first direction is , The first direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the first direction feature differential value of the first feature pixel and the second feature pixel of the second feature pixel are the same. According to determination that the difference between the absolute value of the one-way characteristic differential value is equal to or less than a first predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the second direction line data in which a value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set. ,
In the step of outputting the second direction line data,
In the second direction, the second direction feature differential value of the third feature pixel that is the feature pixel existing within the first predetermined range related to the thickness of the line from the position of the first feature pixel is , The second direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the second direction feature differential value of the first feature pixel and the third feature pixel of the third feature pixel are the same. According to the determination that the difference between the absolute value of the two-way characteristic differential value is equal to or less than a second predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the first direction line data in which a value indicating that the first feature pixel and the third feature pixel are pixels constituting an edge of the line in the first direction is set. , Line extraction processing program. A computer-readable recording medium storing the line extraction program according to claim 8. A computer-readable recording medium storing the line extraction program according to claim 9. A method of extracting data corresponding to a line included in image data from image data corresponding to a coordinate plane extending by a coordinate axis extending in a first direction and a coordinate axis extending in a second direction orthogonal to the axis. ,
Inputting the image data;
Differentiating the image data in a first direction to create first direction differential value data;
Generating first direction feature differential value data in which the first direction differential value having the maximum absolute value in a positive or negative continuous section of the first direction differential value data is used as the first direction feature differential value. And steps to
Outputting second direction line data representing information of a line corresponding to the second direction,
The feature pixel taking the first direction differential value includes a first feature pixel whose first direction feature differential value is other than 0,
In the step of outputting the second direction line data,
The first direction feature differential value of the second feature pixel which is the feature pixel existing within the first predetermined range regarding the thickness of the line from the position of the first feature pixel in the first direction is , The first direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the first direction feature differential value of the first feature pixel and the second feature pixel of the second feature pixel are the same. According to determination that the difference between the absolute value of the one-way characteristic differential value is equal to or less than a first predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the second direction line data in which a value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set. , Line extraction processing method. A method of extracting data corresponding to a line included in image data from image data corresponding to a coordinate plane extending by a coordinate axis extending in a first direction and a coordinate axis extending in a second direction orthogonal to the axis. ,
Inputting the image data;
Differentiating the image data in a first direction to create first direction differential value data;
Generating first direction feature differential value data in which the first direction differential value having the maximum absolute value in a positive or negative continuous section of the first direction differential value data is used as the first direction feature differential value. And steps to
Outputting second direction line data representing information of a line corresponding to the second direction;
Differentiating the image data in the second direction to generate second direction differential value data;
Generating second direction characteristic differential value data in which the second direction differential value having the maximum absolute value in a positive or negative continuous section of the second direction differential value data is used as the second direction characteristic differential value. And steps to
Outputting first direction line data representing information of a line corresponding to the first direction;
Creating line data by synthesizing the data output by the step of outputting the first direction line data and the step of outputting the second direction line data;
A first direction characteristic differential value that takes the first direction differential value includes a first characteristic pixel other than 0;
In the step of outputting the first direction line data,
The first direction feature differential value of the second feature pixel which is the feature pixel existing within the first predetermined range regarding the thickness of the line from the position of the first feature pixel in the first direction is , The first direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the first direction feature differential value of the first feature pixel and the second feature pixel of the second feature pixel are the same. According to determination that the difference between the absolute value of the one-way characteristic differential value is equal to or less than a first predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the second direction line data in which a value indicating that the first feature pixel and the second feature pixel are pixels constituting an edge of the line in the second direction is set. ,
In the step of outputting the second direction line data,
In the second direction, the second direction feature differential value of the third feature pixel that is the feature pixel existing within the first predetermined range related to the thickness of the line from the position of the first feature pixel is , The second direction feature differential value of the first feature pixel is opposite in sign, and the absolute value of the second direction feature differential value of the first feature pixel and the third feature pixel of the third feature pixel are the same. According to the determination that the difference between the absolute value of the two-way characteristic differential value is equal to or less than a second predetermined value for determining that both pixels are pixels corresponding to two edges on both sides of the line. And outputting the first direction line data in which a value indicating that the first feature pixel and the third feature pixel are pixels constituting an edge of the line in the first direction is set. , Line extraction processing method.

Images