JP4296801B2 - Spatial filter, spatial filter creation method, spatial filter creation method program, and image processing apparatus and inspection apparatus using the spatial filter - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a spatial filter used for image processing, and in particular, a spatial filter creation method for emphasizing and detecting line components and edge components according to angles, an automatic calculation program for creating a spatial filter, and the space The present invention relates to a line and edge component detection method and an inspection apparatus in an image to be inspected using a filter.
[0002]
[Prior art]
For example, line components and edges in various inspections such as detection of linear defects existing on the display screen of the display device in screen inspection after manufacturing the display device, or detection of boundary lines between parts and spaces in the appearance inspection of components A method using image processing is used when detecting a (contour) component (hereinafter simply referred to as line detection, and a line includes a line segment). In this method, an object to be inspected is imaged by an imaging means such as a CCD camera, and the obtained image data is subjected to image processing using a spatial filter (hereinafter simply referred to as a filter), and line components in the image data or Detection is performed by enhancing the edge component. Image processing using a filter is to determine a new gradation value by performing a certain local operation on the gradation value of an area including neighboring pixels at any pixel on the image, By performing this processing on all the pixels, enhancement processing, noise removal processing, and the like are performed. Specifically, the filter has a component that is a coefficient set so as to emphasize a detection target by performing a convolution operation for each gradation value of each pixel in a region including a neighboring pixel centered on the target pixel. Represented by a dimensional matrix. As a filter used for line detection, for example, a Sobel filter, a Robinson filter, a Prewitt filter, a Kirsch filter, and the like are known (see Non-Patent Document 1). These filters are filters composed of integer values of 3 × 3 or 5 × 5.
[0003]
[Non-Patent Document 1]
"Image Analysis Handbook", Mikio Takagi, Yoshihisa Shimoda, The University of Tokyo Press, (1991).
[0004]
[Problems to be solved by the invention]
When line detection is performed from image data to be inspected using the above-described conventional filters as they are, processing may be performed using these filters only for the purpose of detecting line components from the image. However, if you want to emphasize only the line component that forms an angle from the image data of the object to be inspected, or if you want to change the processing depending on the angle of the line component, or if you want to change the threshold for making detection judgment based on the angle of the line component For example, a problem occurs when processing that accurately considers the angle component is performed. For example, when a 3 × 3 filter is used, it can be classified only into components in eight directions. That is, detection is performed with an accuracy of 45 °, and the line components at various angles in the image data to be inspected emphasized by the filter are 0 °, 45 °, 90 °, 135 °, It can be classified and included in any result with emphasis on the line components forming 180 °, 225 °, 270 ° and 315 °, but it can be classified into line components having other angles with higher accuracy. It cannot be done and effective processing cannot be performed.
[0005]
In order to improve this, a 5 × 5 filter or a filter of a size larger than that has been proposed. However, since each component of the filter is composed of integer values, an accurate angular component is considered. In addition, the points of each component of the filter take discrete numerical values, and problems such as detection sensitivity changing depending on the angle also occur. In addition to the above-described filters, a second-order differential filter (second-order difference) filter such as a Laplacian filter is used for the purpose of detecting line components. With such a filter, a portion where the intensity change of the image is strong is detected. The angle component is not considered.
[0006]
Therefore, the present invention has been made to solve such problems, and an object thereof is to obtain a filter creation method and the like that can more accurately detect components including up to angular components. . It is another object of the present invention to obtain a method for creating a new image after performing conversion by applying the method.
[0007]
[Means for Solving the Problems]
The filter creation method used for image processing according to the present invention includes a step of setting a region for each component constituting a reference spatial filter, and each component constituting a new spatial filter created from the reference spatial filter. The process of setting the area and the center of the reference spatial filter as the axis, the reference spatial filter is rotated at an angle, and the reference spatial filter area after rotation is divided by the new spatial filter area Then, for each component constituting the reference spatial filter after rotation, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided into divided areas. And for each region of each component constituting the new spatial filter, add the value of the divided region after rotation of the reference spatial filter included in the region, And a step of the value of each component constituting the Tana filter.
In the present invention, for example, in order to create a filter that is most emphasized when forming a certain angle when the shape of the detection target exists at various angles in the image data of the inspection target, The reference filter that emphasizes the detection target at the reference angle (which will also define the angle reference) is rotated by the desired angle, projected onto the new filter coordinate system, and the rotated reference filter Calculate how each component area is divided by each component area constituting a new filter as a percentage of area, and configure each reference filter after rotation based on that ratio The numerical values of the components are divided, and for each region of each component constituting the new filter, the values obtained by the division existing in that region are added to obtain the value of each component constituting the new filter. Therefore, based on the reference filter that detects the shape of the reference angle, a filter that emphasizes the shape forming an angle can be created by calculation processing. Since only the numerical values used in the filter are used, the difference in detection sensitivity depending on the angle can be reduced. In addition, since the value of each component constituting the new filter is constituted by a real value instead of an integer value, the difference in detection sensitivity due to the angle can be further reduced.
[0008]
In addition, the filter creation method according to the present invention uses the values of the components constituting the reference filter as real numbers.
In the present invention, in addition to using an existing integer filter as a reference filter, the value of each component constituting the reference filter is a real number so that more optimal detection can be performed. May be used.
[0009]
In addition, the filter creation method according to the present invention uses the value of each component constituting a new spatial filter as a real number.
In the present invention, the value of each component constituting the new filter is configured by a real number so that more optimal detection can be performed. Thereby, the difference of the detection sensitivity by an angle about the created new filter can be made still smaller.
[0010]
In the filter creation method according to the present invention, the area set for each component constituting the filter is set as a square having the same area.
In the present invention, the area set for each component is set as a square having the same area so as to improve the efficiency of subsequent calculations. By setting in this way, the area ratio when applied from 0 ° to 90 ° is stored, and even when applied to an angle of 90 ° or more, the data of the remainder value divided by 90 is read. It can be applied.
[0011]
In the filter creation method according to the present invention, the reference filter is a filter for enhancing a line component or an edge component.
In the present invention, since a filter that emphasizes a line component is used as a reference filter, effective detection can be performed by using a filter created by the filter creation method according to the present invention for an angle-dependent line component. It can be performed.
[0012]
The filter used for image processing according to the present invention sets a region for each component constituting a reference filter, and sets a region for each component that constitutes a new filter created from the reference filter. With the center of the reference filter as the axis, the reference filter is rotated at a certain angle, the reference filter area after rotation is divided into new filter areas, and the reference filter after rotation is divided. The ratio of the area of the divided region is calculated for each component that constitutes the component, and the value of the divided region is calculated by dividing the value of each component of the reference filter based on the proportion, and a new filter is configured. For each component region to be added, the value of the divided region after rotation of the reference filter included in the region is added to obtain a component constituting a new filter.
In the present invention, the numerical value of the component is effective for each component constituting the filter, which is usually a value at a point (point), with respect to the reference filter (which also defines the angle reference). Temporarily set an area that can be considered. At the same time, for each component constituting the newly created filter, an area in which the numerical value of the component can be regarded as valid is temporarily set. Then, the reference filter is rotated at a certain angle, projected onto the area constituting the new filter, and the ratio of the area of the area divided by the area of each component constituting the new filter is calculated. Based on the ratio, the value of each component of the reference filter is divided to calculate the value of the divided region, and the value of the divided region included in that region is added for each region of each component constituting the new filter. The value is a component. Therefore, it is possible to create a filter that emphasizes a shape forming an angle. In addition, the difference in detection sensitivity due to the angle can be further reduced by expressing the values of the components constituting the new filter by real values instead of integer values.
[0013]
The program for creating a filter used for image processing according to the present invention includes a function for setting a region for each component constituting a reference filter and each component constituting a new filter created from the reference filter. The function to set the area for the center and the center of the reference spatial filter as an axis, the reference spatial filter is rotated at a certain angle, and the reference spatial filter area after rotation is changed by the new spatial filter area. Divide and calculate the ratio of the area of the divided area for each component constituting the reference spatial filter after rotation, and divide the value of each component of the reference spatial filter based on that ratio For each component area that constitutes a new spatial filter, the value of the divided area after rotation of the reference spatial filter included in that area The processing means performs the function of adding and making the value of each component constituting the new filter.
In the present invention, for example, in order to cause the processing means to create a filter that is most emphasized when a desired shape forms an angle, a reference filter (which also defines an angle reference) is provided. When the image is rotated by a certain angle and projected onto the coordinate system before the rotation, how each component area constituting the rotated filter is divided and included in each component area constituting the pre-rotation filter Calculate the ratio of the area, divide the numerical value of each component constituting the rotated filter based on that ratio, and divide the value of the divided area obtained by the division for each area of each component constituting the new filter The values are added to the value of each component constituting the new filter. Therefore, it is possible to create a filter that emphasizes a shape forming an angle. In addition, the difference in detection sensitivity due to the angle can be further reduced by expressing the values of the components constituting the new filter by real values instead of integer values.
[0014]
An image processing apparatus according to the present invention Set the area for each component that makes up the reference spatial filter, set the area for each component that makes up the new spatial filter created from the reference spatial filter, and then set the center of the reference spatial filter For each component constituting the reference spatial filter after rotation by rotating the reference spatial filter at a certain angle about the axis, dividing the reference spatial filter area after rotation by the new spatial filter area Then, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided to calculate the value of the divided area, and each component constituting the new spatial filter For each area, the value of the divided area after rotation of the reference spatial filter included in the area is added to obtain the value of each component constituting the new filter. And a filter, And at least processing means for performing image processing based on the spatial filter.
In the present invention, the image processing apparatus performs image processing using the created filter, and performs detection according to the angle component. Therefore, it is possible to obtain an image processing apparatus capable of performing processing for enhancing a shape forming an angle.
[0015]
Moreover, the inspection apparatus according to the present invention includes: Set the area for each component that makes up the reference spatial filter, set the area for each component that makes up the new spatial filter created from the reference spatial filter, and then set the center of the reference spatial filter For each component constituting the reference spatial filter after rotation by rotating the reference spatial filter at a certain angle about the axis, dividing the reference spatial filter area after rotation by the new spatial filter area Then, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided to calculate the value of the divided area, and each component constituting the new spatial filter For each area, the value of the divided area after rotation of the reference spatial filter included in the area is added to obtain the value of each component constituting the new filter. And a filter, At least a detection unit that detects a line component or an edge component from image data based on a spatial filter, and an imaging unit that captures an image of an inspection target and transmits a signal including image data to the line detection unit. is there.
In the present invention, based on the image data of the inspection object imaged by the imaging unit, the detection unit detects a line component or an edge component from the image represented by the image data using the filter. Therefore, it is possible to inspect by inspecting a line forming an angle and detecting the shape of the inspection target or detecting a line defect or the like in the display screen.
[0016]
In addition, for example, when creating a new image obtained by converting a certain image in order to perform image enlargement / reduction processing, an area is set for each pixel constituting the reference original image and conversion is performed. After that, the image is projected onto the area set for each pixel constituting the new image, the numerical value of each pixel is divided based on the ratio of the converted area divided by the area, and the obtained value is newly set. It adds for every area | region of each pixel which comprises a new image, and it is set as the value based on each pixel which comprises a new image. Therefore, an image corresponding to the conversion process (enlargement / reduction, etc.) can be created.
[0017]
DETAILED DESCRIPTION OF THE INVENTION
Embodiment 1. FIG.
FIG. 1 is a block diagram showing a configuration of an inspection system according to the first embodiment of the present invention. In FIG. 1, an image processing inspection apparatus 1 creates a filter for emphasizing and detecting a line according to an angle formed by a line included in an image, and uses the line detection filter to image an object to be inspected. Data line detection processing is performed and inspection is performed. The image processing inspection apparatus 1 further includes a filter creation processing unit 1A, a filter calculation processing unit 1B, and a display processing unit 1C. The filter creation processing unit 1A performs a process of creating a new filter corresponding to the angle component based on a reference filter prepared in advance (hereinafter referred to as a reference filter). The filter calculation processing means 1B performs calculation (image processing) using the filter created by the filter creation processing means 1A. The display processing unit 1C transmits a display signal for displaying the result calculated by the filter calculation processing unit 1B, for example, as an image.
[0018]
The display device 2 displays the calculation result of the filter processing and the inspection result based on the display signal from the display processing means 1C. The instruction input device 3 is an input device represented by a pointing device such as a keyboard or a mouse, for example, and transmits instructions, data, and the like input by the operator to the image processing inspection apparatus 1 as input signals. The storage device 4 temporarily or long-term stores data processed by the image processing inspection apparatus 1. The imaging device 5 is a device that captures an image of an object to be inspected, and includes, for example, a CCD camera having a plurality of light receiving elements such as a CCD. Each light receiving element converts a received light amount into, for example, an electrical signal, Transmit as a video signal. Reference numeral 100 denotes a display panel to be inspected.
[0019]
In this embodiment, a filter that appropriately detects a line component that forms a desired angle is created, and a line component that forms a desired angle is detected from an object to be inspected by image processing using the filter, and inspection is performed. is there. Therefore, first, a new filter that appropriately detects a desired angle is created from the line detection filter corresponding to the reference angle. Here, in the present embodiment, for each filter component that has been conventionally performed with an integer value, at least a newly created filter component is configured with a real number (including a case where the result is an integer value). In the present embodiment, an area (range) in which the numerical value of each component extends is set for a reference filter (hereinafter referred to as a reference filter) and a newly created filter. Then, the reference filter is rotated by a desired angle, projected onto the area of the new filter coordinate system, and the area ratio of each area of the rotated reference filter depends on each area of the new filter coordinate system. Calculate whether to divide. After dividing the numerical value of each component of the reference filter at that ratio, the value obtained by adding the numerical values of the divided areas of the reference filter existing in each area of the new filter coordinate system is used as the new filter area (or The value of a certain point in the area) and the values of components constituting a new filter. If image processing is performed based on the filter thus calculated, a line component that forms a desired angle from the reference angle can be emphasized and detected. Note that the rotation, projection, and the like of the filter described below are not performed in the actual operation, but are conceptually performed on the program of the image processing inspector apparatus 1.
[0020]
Here, each means constituting the image processing inspection apparatus 1 may be configured as separate means (hardware), but in the present embodiment, the image processing inspection apparatus 1 is replaced with, for example, a CPU (Central Processing Unit). Assume that the computer is composed of arithmetic processing means such as a central computer. Then, the processing procedure of each unit is programmed in advance, the arithmetic processing unit executes the processing based on the program, thereby realizing the function of each unit described above.
[0021]
FIG. 2 is a diagram illustrating an example of a line having an arbitrary angle. In the present embodiment, a filter that emphasizes a line component in the vertical (vertical) direction is defined as a filter that emphasizes a line having an angle of 0 °. The direction in which this line is turned clockwise is set as the plus direction. In this embodiment, the setting is made as described above. However, if the same processing system is used, there is no particular problem even if, for example, the horizontal (horizontal) direction is 0 ° or any other angle is 0 °. Further, the plus / minus of the direction of rotation is not limited, and the counterclockwise direction may be set in the plus direction.
[0022]
Next, a method for creating a filter according to the present embodiment will be described. FIG. 3 is a diagram illustrating a reference filter used in the present embodiment. Here, in the present embodiment, it is assumed that a real number (floating point number) is used as the value of each component constituting the filter. The filter of FIG. 3 which is a reference filter is one in which the value of each component constituting the filter is set so that a line in the 0 ° direction having a width of 3 to 4 pixels in the horizontal (horizontal) direction can be mainly detected. When creating the reference filter, for example, an optimum value of each component is calculated and set with a real value so as to detect a line in the direction of 0 ° by a filter creation program using a genetic algorithm. Here, the component in the row direction of the reference filter that detects the line component in the 0 ° direction (FIG. 3A) constitutes an optimal numerical value for emphasizing the above-described line component, and the column direction The components in the row direction are copied and arranged for the length of the line to be detected. The filter configured as described above can mainly emphasize lines longer than the length in the column direction. Therefore, as the number of components in the column direction of the filter increases, the range of the angle of the emphasized line is narrowed by selectively reacting to the desired angle line (in this case, 0 °). And if the number of components in the column direction of the filter decreases, even if it is not a line component, it has meandered even if it has a feature that is emphasized by the filter in the image, for example, unevenness Since the line can also be emphasized, the angle of the emphasized line is increased. Therefore, the filter that selectively emphasizes the angle component requires a certain length in the column direction. Here, FIG. 3B includes nine columns, and can mainly detect vertical lines with a length of 9 pixels or more.
[0023]
FIG. 4 is a conceptual diagram showing a case where the reference filter is rotated to an angle to be detected. A calculation procedure for creating a filter will be described with reference to FIG. Here, image processing generally performed by a processing device such as a computer handles discrete data. For example, numerical values constituting image data used in image processing are values corresponding to each point of the image sensor, and usually do not include area (region) information. The same applies to the filter. In the present embodiment, this is conceptually regarded as having an area where each component of the filter is effective. The reference filter used in the present embodiment is assumed to be composed of a 9 × 9 square area, and is assumed to have a value for each component in each area. Each area is represented by a two-dimensional coordinate value (the horizontal direction is the X direction and the vertical direction is the Y direction).
[0024]
First, the angle of the line component to be detected by the newly created filter is determined. Here, a line component of about 5 ° is detected. Then, the reference filter is rotated at that angle as shown in FIG. In FIG. 4, a reference filter after rotation (hereinafter referred to as a rotation filter) is represented by a dotted line. A solid line is area information constituting a new filter. In the present embodiment, the same information as the coordinate system of the reference filter before rotation is used. Here, the size of the new filter is larger than the size of the reference filter before the rotation. This is because the filter newly created by rotating the filter has a maximum of about 1 compared to the reference filter. Because the size is 4 times larger (in the case of the 9 × 9 standard filter, the maximum filter size is 13 × 13 as in this embodiment), the size of the new filter is Prepare for the minutes. Actually, the coordinate system (coordinates) of a new filter to be prepared is limited due to the saving of hardware resources such as memory, but it is better to provide a margin as much as possible.
[0025]
FIG. 5 is an enlarged view of a part of FIG. In FIG. 5, the portion surrounded by a circle in FIG. 4 (position near the area (coordinates) (8, 3) of the rotated reference filter) is enlarged. As shown in FIG. 5A, the numerical value of the (8, 3) component of the reference filter is −7.68. Next, conceptually projecting the rotated reference filter area onto a new filter coordinate system (hereinafter referred to as a new coordinate system), each component of the rotated reference filter is configured in the new coordinate system. Divided by areas. A ratio of the area of the divided area (hereinafter referred to as an area ratio) is calculated (FIG. 5B). The (8,3) component of the rotated reference filter is divided into four parts by area ratios of 60%, 20%, 15%, and 5%, respectively, according to the new filter area. The filter creation means 1A further divides the reference filter component values by the calculated area ratio (FIG. 5C). Here, since the value of the component of the reference filter is −7.68, when the area ratio is multiplied, −4.61, −1.54, −1.15, and −0.38 from the descending order of the area of the divided areas. It becomes.
[0026]
FIG. 6 is a conceptual diagram illustrating a numerical value calculation method for new filter components. Division of each component of the reference filter as described above is performed for all areas of the reference filter, and in each area configured with a new coordinate system, the calculated numerical values of the divided areas included in the area are calculated. Each is added (FIG. 6A). The numerical value calculated in this way becomes a value of a new filter component that appropriately detects a line component that forms a desired angle to be detected (FIG. 6B).
[0027]
FIG. 7 is a diagram showing a filter created by rotating the reference filter by 15 ° so as to create a filter for detecting a 15 ° line component. Here, the value of each component of the filter of FIG. 7 is calculated using the Monte Carlo method which is a general algorithm, and can be divided by the rotated reference filter and each area created in the new coordinate system. After calculating the area ratio of the area using a random number, the value of each component is calculated. Although the Monte Carlo method is used in the present embodiment, the method for calculating the area ratio of the divided areas is not limited to this, and for example, the area may be calculated by calculating the area based on the seating calculation for all areas. good. The filter shown in FIG. 7 is actually a 13 × 13 size filter, but when rounding off three decimal places that are considered to have little effect on the filter processing calculation results, the outermost values are all 0. Therefore, here, the filter has a size of 11 × 11.
[0028]
The filter calculation processing means 1B performs filter processing on the image data to be inspected using the new filter created by the filter creation processing means 1A as described above, and emphasizes and detects line components at a desired angle. To do.
[0029]
8, 9 and 10, three types of filters having different detection angles are created by the filter creation processing unit 1 </ b> A, and the same filter to be inspected is used by the filter computation processing unit 1 </ b> B using the created filter. It is a figure showing the result of having performed the filter calculation process with respect to image data, and having detected the line component. 8 is a filter prepared by rotating the reference filter by 90 °, FIG. 9 is a filter prepared by rotating the reference filter by 105 °, and FIG. 10 is a filter processing using a filter prepared by rotating the reference filter by 120 °. The result of having performed. For example, the display processing unit 1C uses the filter created by the filter creation processing unit 1A by the filter computation processing unit 1B to obtain image data obtained by performing filter computation processing on the image data to be inspected, as shown in FIG. 9 and 10 are displayed. As shown in FIGS. 8, 9, and 10, a line having an angle component of 105 ° or the vicinity thereof, which is hardly detected in FIGS. 8 and 10, is detected toward the upper right of the display screen of FIG. You can see that
[0030]
As described above, according to the first embodiment, a region is conceptually set for each component of the reference filter and the new filter to be created, and then the reference filter is rotated by a desired angle to obtain new coordinates. Projected onto the system, the area of the reference filter after rotation is divided by the area of the new coordinate system, and the area ratio is determined by determining what area ratio is included in the area of the new coordinate system. Accordingly, the value of the component of the reference filter is divided, and the value of the divided area included in each area of the new coordinate system is added to create a filter with the value of the new filter component. A filter that emphasizes the shape that forms the desired angle can be created, and a filter that detects any angle is created based on the same reference filter and using only its constituent values. And for that, it is possible to reduce the difference in detection sensitivity depending on the angle. In addition, since the value of each component constituting the new filter is a real value, the difference in processing sensitivity (detection sensitivity) depending on the angle can be further reduced. And about the inspection apparatus which detected using such a filter, since a line component can be selectively emphasized according to an angle, the setting of a threshold value etc. must be changed according to an angle component. It can correspond to processing.
[0031]
Embodiment 2. FIG.
In the first embodiment described above, as shown in FIG. 2, the values of the components constituting the reference filter are floating point numbers. This invention is not limited to this, For example, you may use the existing filter whose component which comprises a reference | standard filter is an integer value. Of course, even if the reference filter component is an integer value, the numerical value of the created new filter component is not necessarily an integer, and the numerical value of the new filter component is a real number.
[0032]
Embodiment 3. FIG.
FIG. 11 is a conceptual diagram showing a case where an image is enlarged or reduced. In the above-described embodiment, the area of each component of the reference filter is set as a square having the same area. However, the present invention is not limited to this, and the shape and the like are arbitrarily set depending on the filter to be created. be able to. Although not particularly mentioned in the above-described embodiment, of course, there is no difference from a normal filter except that the numerical value of the component corresponding to the angle is calculated.
[0033]
In the above-described embodiment, a new filter is created by rotating the reference filter, but this can also be applied to image processing for enlarging or reducing an image. For example, when the original image is enlarged or reduced to a desired size, a value (for example, luminance value) based on the pixels constituting the new image is determined by calculating the area ratio included in the area of the new image. Therefore, it can be applied to image conversion including enlargement / reduction.
[0034]
Embodiment 4 FIG.
In the above-described embodiment, the image processing inspection apparatus 1 is configured by the filter creation means 1A that performs new filter creation processing, the filter calculation processing means 1B that performs actual image processing, and the display processing means 1C. Only the filter creation means 1A may be configured by an independent device. Data of a new filter created by another device can be stored in the storage device 4, for example, and image processing can be performed.
[0035]
Embodiment 5. FIG.
In the first embodiment described above, the filter creation unit 1A performs a series of processes for creating a new filter based on the reference filter and the desired angle. For example, an area corresponding to the angle is previously provided to a certain reference filter. The ratio may be calculated and stored. The new filter actually created has different component values depending on the component values of the reference filter, but the divided area ratio does not change in relation to the rotation filter and the grid created in the original coordinate system. Therefore, it can be applied when creating other filters. This can be applied to a reference filter having a size smaller than the stored reference filter. In particular, when the area is defined by a square, if the area ratio up to 90 ° is calculated, the area ratio up to 90 ° can be made to correspond to the angle beyond that. In addition, even when it is defined by a rectangle, the area ratio up to 180 ° may be calculated.
[0036]
Embodiment 6. FIG.
In the above-described embodiment, line detection processing is used for the inspection of the display screen of the display device, but the present invention is not limited to this, and for line detection for other inspections and other uses, The filter creation method, the filter, and the image processing using the filter of the present invention can be applied.
[0037]
In the above-described embodiment, the filter creation method is applied to the line detection filter. However, the present invention is not limited to this. For detection of a shape other than a line, such as an ellipse, that has an angle dependency, the angle The present invention can also be applied to detection in consideration of the above.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an inspection system.
FIG. 2 is a diagram illustrating an example of a line having an arbitrary angle.
FIG. 3 is a diagram illustrating a reference filter used in the present embodiment.
FIG. 4 is a conceptual diagram showing a case where rotation is performed at an angle to be detected.
5 is an enlarged view of a part of FIG.
FIG. 6 is a diagram illustrating numerical calculation of new components.
FIG. 7 is a diagram showing a filter created by rotating 15 °.
FIG. 8 is a diagram illustrating a processing result by a filter created by 90 ° rotation.
FIG. 9 is a diagram illustrating a processing result by a filter created by a 105 ° rotation.
FIG. 10 is a diagram illustrating a processing result by a filter created by rotation by 120 °.
FIG. 11 is a conceptual diagram illustrating a case where an image is enlarged or reduced.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Image processing test | inspection apparatus, 1A filter preparation process means, 1B filter calculation process means, 1C display process means, 2 display apparatus, 3 instruction input device, 4 memory | storage device, 5 imaging device

Claims (10)

In the method of creating a spatial filter used for image processing, a step of setting a region for each component constituting a reference spatial filter;
Setting a region for each component constituting a new spatial filter created from a reference spatial filter;
Rotate the reference spatial filter at a certain angle around the center of the reference spatial filter, divide the rotated reference spatial filter area by the new spatial filter area, Calculating the ratio of the area of the divided region for each component constituting the spatial filter, and dividing the value of each component of the reference spatial filter based on the ratio to obtain the value of the divided region;
For each component area constituting the new spatial filter, adding the value of the divided area after rotation of the reference spatial filter included in the area to obtain the value of each component constituting the new filter; A spatial filter creating method characterized by comprising:   2. The spatial filter creation method according to claim 1, wherein the value of each component constituting the reference spatial filter is a real number.   The spatial filter creation method according to claim 1 or 2, wherein a value of each component constituting the new spatial filter is a real number.   4. The spatial filter creation method according to claim 1, wherein a region set for each component constituting the spatial filter is set as a square having the same area.   The spatial filter creation method according to claim 1, wherein the reference spatial filter is a spatial filter for enhancing a line component or an edge component.   The ratio of the area of the divided area of the reference spatial filter divided by the new area of the spatial filter for each angle when the reference spatial filter is rotated at a certain angle is stored in the storage means. The spatial filter creation method according to claim 1, wherein:   In the spatial filter used for image processing, after setting a region for each component constituting the reference spatial filter, and after setting a region for each component constituting a new spatial filter created from the reference spatial filter, Rotate the reference spatial filter at a certain angle around the center of the reference spatial filter, divide the rotated reference spatial filter area by the new spatial filter area, For each component constituting the spatial filter, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided to calculate the value of the divided area. For each component area that makes up a spatial filter, the value of the divided area after rotation of the reference spatial filter included in that area is added to form a new filter. Spatial filter, characterized in that the value of each component. In a program that creates a spatial filter for image processing,
A function for setting a region for each component constituting the reference spatial filter;
A function to set a region for each component constituting a new spatial filter created from a reference spatial filter;
Rotate the reference spatial filter at a certain angle around the center of the reference spatial filter, divide the rotated reference spatial filter area by the new spatial filter area, A function of calculating the area ratio of the divided area for each component constituting the spatial filter, and dividing the value of each component of the spatial filter serving as a reference based on the ratio to obtain the value of the divided area;
For each component region that constitutes a new spatial filter, the value of the divided region after the rotation of the reference spatial filter included in that region is added, and the value of each component that constitutes the new filter is obtained. A program for a spatial filter creation method characterized by causing a processing means to perform. Set the area for each component that makes up the reference spatial filter, set the area for each component that makes up the new spatial filter created from the reference spatial filter, and then set the center of the reference spatial filter For each component constituting the reference spatial filter after rotation by rotating the reference spatial filter at a certain angle about the axis, dividing the reference spatial filter area after rotation by the new spatial filter area Then, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided to calculate the value of the divided area, and each component constituting the new spatial filter For each area, the value of the divided area after rotation of the reference spatial filter included in the area is added to obtain the value of each component constituting the new filter. And a filter,
An image processing apparatus comprising at least processing means for performing image processing based on the spatial filter. Set the area for each component that makes up the reference spatial filter, set the area for each component that makes up the new spatial filter created from the reference spatial filter, and then set the center of the reference spatial filter For each component constituting the reference spatial filter after rotation by rotating the reference spatial filter at a certain angle about the axis, dividing the reference spatial filter area after rotation by the new spatial filter area Then, the ratio of the area of the divided area is calculated, and based on the ratio, the value of each component of the reference spatial filter is divided to calculate the value of the divided area, and each component constituting the new spatial filter For each area, the value of the divided area after rotation of the reference spatial filter included in the area is added to obtain the value of each component constituting the new filter. And a filter,
Detection means for detecting line components or edge components from the image data based on the spatial filter;
An inspection apparatus comprising at least imaging means for imaging an object to be inspected and transmitting a signal including image data to the line detection means.

Images