JP7343133B1 - Image inspection methods and programs - Google Patents

Abstract

The present invention provides an image inspection method for extracting only a specific part from an image including arbitrary curved parts in a limited manner and converting the extracted part into a rectangular image. [Solution] An image inspection method for determining the quality of an inspection object using an image obtained by photographing the inspection object includes a first step of setting a required observation part at a desired location of the inspection object in the image. , a second step of acquiring linear or rectangular pixel values of a predetermined length extending in a predetermined direction from each of the consecutive pixels constituting the observation target portion. Preferably, the method further includes a third step of aligning the acquired pixel values into a new rectangular image. [Selection diagram] Figure 1

Description

The present invention relates to an image inspection method and program for determining the quality of an inspection object using images obtained by photographing the inspection object.

Image inspection of industrial products is aimed at detecting defects such as scratches and stains, and generally involves a method of detecting differences from images of normal products.
In recent years, many methods have been used in which, for example, a normal image or a plurality of defective images are trained by deep learning or the like to determine the presence or absence of a defect in an image of an object to be inspected. Furthermore, machine learning methods such as the MT method (see Non-Patent Document 1) are also known, in which only information on a reference image is learned and the presence or absence of defects in a target image is detected. Furthermore, although it is a classical method, a method is also used in which determination is made based on the magnitude of the difference between a binarized image and a normal image. In this specification, these algorithms for determining the presence or absence of defects are collectively referred to as "inspection algorithms."

There are various products that are subject to image inspection, but for example, image inspection of die-cast products for automobile engines is aimed at automating visual inspection by humans, and the inspection accuracy is equivalent to or higher than that of visual inspection. is required.
Die-cast products are manufactured by pouring molten aluminum or other material into a mold and removing it from the mold after the material has cooled. In visual inspection of a product, the presence or absence of burrs, chips, hot water wrinkles, etc. is inspected, and the location of the defect is often determined by the type of defect. For example, burrs occur at mold matching areas such as the outer periphery of die-cast products, and chips frequently occur at corners and ridges of die-cast products, so inspectors should focus their inspections on such areas. . In this way, the inspector's attention area and judgment conditions generally differ depending on the type of defect.
Since the purpose of image inspection is to automate visual inspection by inspectors, it is desirable to set the optimal image location for each defect and pass it to the inspection algorithm when performing image inspection.

Shoichi Teshima et al., “Research on appearance inspection technology applying the Mahalanobis-Taguchi system method,” Quality Engineering, Society for Quality Engineering, October 1997, Volume 5, No. 5, p. 38-45

As mentioned above, the areas to be focused on during inspection differ depending on the type of product defect. In particular, when the object to be inspected has a complicated shape, in order to perform an efficient inspection, it is desirable to focus the inspection on points of interest (for example, ridges, outer periphery). However, in conventional image inspection, since the image taken for inspection is rectangular, there is a problem that the inspection image includes areas other than the area of interest (for example, when trying to inspect a chip on a ridge line). (In this case, the image also includes areas other than the ridgeline and the stand on which the object is placed.)

In addition, most existing inspection algorithms target rectangular images, so if the point of interest includes a curved part, the inspection algorithm is applied using such a rectangular image as is. Inspection processing is performed on parts including parts other than the target part, which causes an increase in processing time and a deterioration in inspection accuracy.

The present invention was developed under such circumstances, and provides an image inspection method and program for extracting only a specific part from an image including an arbitrary curved part in a limited manner and converting the extracted part into a rectangular image. It is an object.

The image received by the inspection algorithm is generally a collection of pixels arranged in a horizontal and vertical grid, with an overall rectangular shape.
The present inventor believes that even if the inspection target area includes an arbitrary curved shape, if it is possible to extract only the part that requires observation along the curve and convert it into a rectangular image, the inspection algorithm can be used in a short time using existing inspection algorithms. The idea was that it would be possible to perform image processing with high precision.
A method called polar expansion is conventionally known as a technique for extracting a non-rectangular portion from an image and transforming it into a rectangle. Polar expansion calculates the center and diameter of a circular shape identified from the target image, then defines a number of straight lines radially connecting the center and outer circumference (circumference) of the circle, and then calculates the linear pixels. This method aligns them and transforms them into a rectangle. In polar expansion, the center of the circle is greatly enlarged, but the outer periphery becomes a rectangular image, making it easier to apply the inspection algorithm in a short time and with high precision.
However, in polar expansion, in image inspection of die-cast products and the like, it is not possible to extract specific areas where defects are likely to occur (for example, the outer periphery, corners, or ridge lines) as a limited rectangular image.
As will be described in detail below, the present inventor proposes a method and a program for extracting only a specific part from an image including an arbitrary curved part in a limited manner and converting the extracted part into a rectangular image.

The image inspection method of determining the quality of the inspection object using an image obtained by photographing the inspection object according to claim 1 includes: a first step of setting a straight line and/or a curved line, and a second step of obtaining a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each continuous pixel constituting the observation target portion. The method is characterized in that the acquired pixel value is set in a direction perpendicular to a tangent to a virtual curve drawn by a moving average of the pixels.

The image inspection method according to claim 2, in which the quality of the inspection object is determined using an image obtained by photographing the inspection object, is such that an observation-required portion is placed at a desired location of the inspection object in the image. a first step of setting a straight line and/or a curved line, and a second step of obtaining a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each continuous pixel constituting the observation target portion. and the acquired pixel value is set in a direction perpendicular to a straight line connecting pixels at preset intervals among the pixels.

The image inspection method of determining the quality of the inspection object using an image obtained by photographing the inspection object, according to a third aspect of the present invention, includes: a first step of setting a straight line and/or a curved line, and a second step of obtaining a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each continuous pixel constituting the observation target portion. and the acquired pixel value is set by translating the observation target portion by a predetermined distance and connecting corresponding pixels before and after the translation. .

According to a fourth aspect of the present invention, there is provided an image inspection program for determining the quality of the inspection object using an image obtained by photographing the inspection object; is set as a straight line and/or a curved line and stored in a storage device, and a linear or rectangular pixel value of a predetermined length extending in a predetermined direction is obtained from each continuous pixel constituting the observation portion. The method is characterized in that the acquired pixel value is set in a direction perpendicular to a tangent to a virtual curve drawn by a moving average of the pixels.

According to a fifth aspect of the present invention, there is provided an image inspection program for determining the quality of the inspection object using an image obtained by photographing the inspection object, which includes an image inspection program for determining the quality of the inspection object using an image obtained by photographing the inspection object. is set as a straight line and/or a curved line and stored in a storage device, and a linear or rectangular pixel value of a predetermined length extending in a predetermined direction is obtained from each continuous pixel constituting the observation portion. acquiring and storing it in a storage device, and the acquired pixel value is set in a direction perpendicular to a straight line connecting pixels at preset intervals among the pixels. .

According to a sixth aspect of the present invention, there is provided an image inspection program for determining the quality of the inspection object using an image obtained by photographing the inspection object; is set as a straight line and/or a curved line and stored in a storage device, and a linear or rectangular pixel value of a predetermined length extending in a predetermined direction is obtained from each continuous pixel constituting the observation portion. the acquired pixel value is set by translating the observation-required portion by a predetermined distance and connecting corresponding pixels before and after the translation. It is characterized by this.

According to the image inspection method and program of the present invention, it is possible to extract only a specific part from an image including an arbitrary curved part and convert it into a rectangular image in a short time using existing inspection algorithms. It becomes possible to perform image processing with high precision.

1 is a flow diagram of an image inspection method according to an embodiment of the present invention. FIG. 2(a) is a diagram showing a photographed image of a product, and FIG. 2(b) is a diagram showing various defects that may occur in the product. It is a diagram showing the parts of the product that need to be observed. FIG. 4(a) is a diagram showing a portion of a product requiring observation, and FIG. 4(b) is a diagram showing the portion requiring observation as pixels. 5(a) is an enlarged view of the area 5a in FIG. 4(b), FIG. 5(b) is a table showing the coordinate values of pixel <A>, and FIG. 5(c) is a table showing the coordinate values of pixel <D>. The diagram shown in FIG. 5(d) is a diagram showing a straight line <E>. Figure 6 (a) is a diagram showing the straight line <to>, Figure 6 (b) is a diagram showing the straight line <to> in pixels, and Figure 6 (c) is a diagram showing the alignment of pixels including the straight line <to>. This is a diagram. Figure 7(a) is a diagram showing the straight line <li>, Figure 7(b) is a diagram showing the straight line <li> in pixels, and Figure 7(c) is a diagram showing the alignment of pixels including the straight line <li>. This is a diagram. FIG. 7 is a diagram showing the results of executing the second step for all pixels of the observation-required portion. FIG. 9(a) is a diagram in which pixels representing straight lines are arranged in a rectangular shape, and FIG. 9(b) is a diagram in which pixels are shown in FIG. 9(a). 10(a) shows the case where there is no chipping on the ridge line, and FIG. 10(b) shows the case where there is a chip on the ridge line. FIG. FIG. 11(a) is a diagram showing an image taken from directly above the product, and FIG. 11(b) is a diagram showing the result of implementing the image inspection method of the present invention on FIG. 11(a). be. FIG. 7 is a diagram showing another method for finding a tangent. FIG. 13(a) is a diagram showing yet another method for determining a straight line of a predetermined length, and FIG. 13(b) is a diagram for explaining a case where the method of FIG. 13(a) is not applicable. 1A and 1B are various diagrams showing results obtained by the image inspection method of the present invention. FIG. 7 is a diagram showing a method of setting a rectangle instead of a vertical line in the second step.

Next, an image inspection method according to a preferred embodiment of the present invention will be described with reference to the drawings, taking image inspection of a die-cast product (hereinafter referred to as "product") as an example. FIG. 1 is a flow diagram of an image inspection method according to a preferred embodiment of the present invention.

FIG. 2(a) is a diagram showing a photographed image of the product 10. The background of the photographed image is usually black, and in this example, since there is illumination above the product 10, the background is displayed in dark gray, the top surface of the product 10 is displayed in white, and the side surface of the product 10 is displayed in light gray. When manufacturing a product, a male die and a female die (neither shown) are provided at the upper and lower sides, respectively, which are clamped together under high pressure. A molten material is poured into the space between the male mold and the female mold, and after cooling the material, the hardened material is taken out from the mold to obtain the product 10. Therefore, the mold matching portion 10a appears slightly.

FIG. 2(b) is a diagram showing burrs 12, chips 14 on the ridge line portion 10b, and hot water wrinkles 16 on the flat portion that occur during manufacturing of the product 10. Among these defects, burrs 12 are likely to occur at the mold matching portion 10a when the tightening pressure between the male and female molds is insufficient or when the male and/or female molds are thermally deformed. Moreover, the chipping 14 is likely to occur at the ridgeline portion 10b or corner portion 10c of the product 10, and occurs when the injection pressure of the molten material (hereinafter referred to as "hot water") to be poured is insufficient. Furthermore, hot water wrinkles 16 are defects in which unevenness is formed in a place that should be flat, and places where the flow of hot water stagnates (for example, near the gate where hot water flows into the mold, or when hot water is poured into the mold from multiple gates). It is likely to occur at the confluence of hot water.

Now, let us consider a case where the photographed image of FIG. 2(a) is obtained and the presence or absence of a chip 14 in the ridgeline portion 10b visible in the foreground is to be inspected. First, attention will be paid to only the region near the ridgeline portion 10b (the broken line portion in FIG. 3) in the image, and a procedure for converting this region into a rectangular image will be described.

(1st step)
First, as a first step, as shown in FIG. 4(a), a required observation area A is set by a straight line and/or a curved line at a location near the ridgeline 10b that the engineer in charge considers to be inspected. Any method may be used for setting, such as drawing directly on the image with a mouse, touch pen, or the like.

FIG. 4(b) shows the observation-required portion A set in this way. The observation-required portion A is a set of continuous pixels, and in this example, has a downwardly convex curved shape as a whole. Consider rectangular coordinates with the origin at the upper left corner of the observation area A. This takes into account that in image data, the upper left end is generally the origin, the right direction is the positive X axis, and the downward direction is the positive Y axis.

Symbols will be defined for the explanation of the drawings that follow.
Pixel number: i (in this example, i = 0 to 63; there are 2 pixels vertically at X = 62)
Total number of pixels: N (in this example, N=64)
Pixel coordinates: (X _i , Y _i )
Moving average parameter: p

The moving average parameter p is a parameter for calculating a new coordinate value, which will be described later, and the moving average for the pixel number i of interest is given by the following formula.
X'=(X _i-p +X _i-p+1 +...+X _i +X _i+1 +...+X _i+p )/(2p+1)
Y'=(Y _i-p +Y _i-p+1 +...+Y _i +Y _i+1 +...+Y _i+p )/(2p+1)
That is, at the pixel number i of interest on the observation target portion A, the average coordinates of p×2 coordinates on both sides of the pixel and the coordinates of the pixel number i itself are determined.

(Second step)
The second step surrounded by a broken line in FIG. 1 will be explained. FIG. 5(a) shows an enlarged view of the region 5a indicated by the dashed line in FIG. 4(b). First, let's focus on pixel <A>. The coordinates of pixel <a> are (31, 21). When the moving average parameter p is 2, the coordinates including two adjacent pixels are as shown in FIG. 5(b). That is, the coordinate average value of pixel <A> is (31, 20.6), and its position is indicated by a circle <B> in FIG. 5(a).

Next, pay attention to the pixel <C> located to the right of the pixel <A>. The coordinates of pixel <c> are (32,21). In the same manner as in the case of pixel <A>, the coordinate average value (32, 20.8) of pixel <C> is determined, and its position is indicated by a circle <D> in FIG. 5(c). Then, set a straight line <ho> connecting the center of <b> and the center of <d> (Fig. 5(d)). The straight line <E> can be regarded as a tangent to the observation target portion A at the pixel <A>. Note that the reason why we connected the centers of the coordinate average values of adjacent pixels (<B> and <D>) rather than the centers of adjacent pixels (<A> and <C>) is to find a tangent line that is as natural as possible. It is.

Next, as shown in FIG. 6A, a straight line <to> with a length L extending from the center of the pixel <A> in a direction substantially perpendicular to the straight line <E> is set. FIG. 6(b) shows the straight line <to> using pixels. In the example shown in this figure, the straight line <to> is a set including nine pixels including pixel <a> (FIG. 6(b)). In the example shown in FIG. 6(a), the straight line <to> is set below the pixel <A> (diagonally downward), but the setting direction of the straight line <to> is set to obtain a rectangular image. The direction is as follows. That is, in this example, the inspection target is the ridgeline part 10b of the product 10, and since we are trying to obtain a rectangular image of the part located below the observation required part A in FIG. 4(b), the straight line <to> , is set below the pixel <A>, not above it.

Regarding the pixel <C> located on the right side of the pixel <A>, in the same way as the straight line <To> was obtained for the pixel <A>, a straight line <C> and a line extending approximately perpendicular to the straight line <C> are obtained. A straight line <li> of length L is set (FIG. 7(a)). To explain in more detail, new coordinates are determined from the coordinate average value of the pixel <C> and are indicated by a circle <D>. Next, new coordinates are determined from the coordinate average values of the pixel <nu> located on the right side of the pixel <c>, and are indicated by a circle <g>. Then, set a straight line <C> connecting the center of <D> and <G>, and a straight line <R> of length L extending from the center of pixel <C> in a direction almost perpendicular to the straight line <C>. . The straight line <li>, like the straight line <to>, is a set including nine pixels including the pixel <c> (FIG. 7(b)).

The above-described processing performed for pixel <A> and pixel <C> is performed for all pixels of observation target portion A. In this way, the brightness information (hereinafter referred to as "pixel value") possessed by the pixels of the observation target portion A is obtained. The results are shown in FIG. In FIG. 8, only the straight line with the determined length L is displayed. Note that for p=2 pixels located at both ends of the observation-required portion A, the moving average coordinates cannot be calculated, so a straight line of length L is not determined.

(3rd step)
Next, as a third step, pixels including straight lines of length L are arranged into a rectangle. FIG. 6(c) is a diagram in which pixels including a straight line <to> of length L (9 pixels in this case) determined for pixel <a> are aligned; It is a diagram in which pixels including a straight line <i> of length L (also 9 pixels) determined for <c> are arranged. Here, "aligning" means rearranging the pixels in a line. In this way, the diagonally extending straight lines obtained for the pixel <A> and the pixel <C> can be rearranged into a row of vertical pixels like <R> and <W>, respectively. When <ru> and <wo> are placed side by side, an image with horizontal and vertical pixels of 2 x 9 is created.

FIG. 9(a) is a diagram showing a state in which all straight lines of length L are aligned, and FIG. 9(b) is a diagram showing these as pixels. The observation-required portion A is drawn with N pixels, and in this example, the moving average parameter p=2, resulting in (N-4) pieces of image data of length L in the horizontal direction.

Through the above processing, the area indicated by the broken line in FIG. 3 has been converted into a rectangular image. For example, if the ridgeline portion 10b in FIG. 2 is normal, the image will be a rectangular image as shown in FIG. 10(a), but if there is a chip 14 in the ridgeline portion 10b in FIG. The image will look like this.

FIG. 11A is a diagram showing an image taken from directly above the product 10 in order to inspect the presence or absence of burrs 12 on the outer periphery of the product 10. As in the case of inspecting the presence or absence of the chip 14 described above, by setting the observation required part at the place that the engineer in charge thinks should be inspected and converting it to a rectangular image, if there is a burr 12, it can be detected as shown in Fig. 11. An image like (b) can be obtained.

Similarly, the presence or absence of hot water wrinkles 16 on the product 10 can be quickly and easily inspected by converting the image into a rectangular image.

In the above description of the embodiment, in order to obtain a straight line of a predetermined length that is substantially perpendicular to the tangent to the observation area A, the moving average coordinates obtained from the observation area A are used, but other suitable methods may also be used. A straight line of a predetermined length may be obtained.

For example, find a straight line <H'> by connecting the centers of p (in this case 2) neighboring pixels of the pixel of interest <A>, and set a straight line extending almost perpendicular to the straight line <H'>. , it may be a straight line of a predetermined length (see FIG. 12).

Alternatively, for example, a straight line of a predetermined length may be created by translating the set observation-required part by a predetermined distance and connecting the corresponding pixels of the original observation-required part and the parallel-transferred observation-required part. Figure 13(a)). This method is effective when the degree of curvature of the portion to be observed is small. This method cannot be applied when the degree of curvature of the observation-required portion is large and the angles at both ends of the observation-required portion are 180 degrees or more (FIG. 13(b)).

FIG. 14 is a diagram showing various results obtained by the image inspection method of the present invention. In other words, Fig. 14(a) shows a case in which a vertical line is drawn outside a downwardly convex curve, and Fig. 14(b) shows a case in which a number of perpendicular lines are drawn to an arbitrary curve, and a case on the vertical line is drawn. A case in which pixel values can be obtained is shown, and FIGS. 14(c) and 14(d) show cases in which vertical lines are drawn on the inside and outside of the diamond outline, respectively.

Furthermore, in the above embodiment, a method of obtaining a straight line of a predetermined length extending in a direction substantially perpendicular to the tangent to the observation-required portion A in the second step is described, but instead of obtaining such a straight line, A method of obtaining a rectangle may also be used. In the example shown in FIG. 15, a rectangle of 5 pixels×length L is drawn with an overlap of one pixel.

The image inspection method of the present invention may be performed using only the first and second steps without using the third step of aligning pixels. Most of the existing inspection algorithms are configured to inspect rectangular images, and in such cases, the third step is useful. This is because there is little need to implement the third step when using the method.

When applying the image inspection method of the present invention, a positional shift may occur in the photographed normal image or the image to be inspected. In such a case, it is necessary to correct the positional shift using a known method. Good. Known methods include, for example, a method in which a mark is set in advance on the object to be inspected to serve as a reference point for positional correction, and a method in which the correlation of brightness between images is determined and the positional correction is performed so that the correlation is maximized. be.

(Image inspection program)
Next, a program for causing a computer to execute the above steps (first step to third step) will be described. The computer on which this program is executed includes a CPU (Central Processing Unit), memory, storage devices such as a hard disk, input devices such as a keyboard, mouse, and touch pen, display devices, and output devices (all of which are interconnected by a bus). (not shown), or a microchip type processing device.

First, the observation-required part inputted by the input device is stored in the memory. Next, the CPU calculates the coordinate average value of two adjacent pixels (for example, pixels <A> and <C> in FIG. 5(c)) among the consecutive pixels constituting the observation area stored in the memory. , stored in memory. A straight line connecting the centers of the coordinate average values calculated in this way (for example, the straight line <H> in FIG. 5C) is set and stored in the memory. Next, a straight line <to> having a length L extending from the center of the pixel <a> in a direction substantially perpendicular to the straight line <h> is set and stored in the memory.

The above-described process is repeated for all consecutive pixels constituting the observation-required portion, and the pixel values of the observation-required portion are acquired. The obtained straight line of length L is stored in memory.

All of the pixels representing the straight line of length L obtained in this way are arranged into a rectangle and stored in memory. Finally, the obtained data is input to the inspection algorithm from an output device or the like.

In the above-described embodiment, the coordinate average value is used to obtain a straight line of length L extending approximately vertically, but similarly to the image inspection method, pixels at preset intervals may be used. Alternatively, a method may be used in which the portion to be observed is moved in parallel by a predetermined distance. Further, if unnecessary, the step of arranging pixels into a rectangle may not be used.

In the above example, the data is stored in memory, but if the amount of data is large, it is stored in a large capacity storage device such as a hard disk.

The present invention is not limited to the embodiments of the invention described above, and various modifications can be made within the scope of the invention described in the claims, and these are also included within the scope of the present invention. Needless to say, it is something that

For example, although the above-described embodiments have been described in relation to image inspection of die-cast products, the present invention can be applied to any inspection that requires visual inspection, such as inspection of semiconductor patterns.

10 Product 10a Mold matching part 10b Ridge line part 10c Corner part 12 Burr 14 Chip 16 Hot water wrinkle A Part to be observed

Claims (6)

An image inspection method for determining the quality of an inspection object using an image obtained by photographing the inspection object, the method comprising:
a first step of setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line;
a second step of acquiring linear or rectangular pixel values of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion ;
An image inspection method characterized in that the acquired pixel value is set in a direction perpendicular to a tangent to a virtual curve drawn by a moving average of the pixels. An image inspection method for determining the quality of an inspection object using an image obtained by photographing the inspection object, the method comprising:
a first step of setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line;
a second step of acquiring linear or rectangular pixel values of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion;
An image inspection method characterized in that the acquired pixel value is set in a direction perpendicular to a straight line connecting pixels at preset intervals among the pixels. An image inspection method for determining the quality of an inspection object using an image obtained by photographing the inspection object, the method comprising:
a first step of setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line;
a second step of acquiring linear or rectangular pixel values of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion;
An image inspection method characterized in that the acquired pixel value is set by translating the observation target portion by a predetermined distance and connecting corresponding pixels before and after the translation. An image inspection program for determining the quality of an inspection object using an image obtained by photographing the inspection object,
setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line and storing it in a storage device;
acquiring a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion and storing it in a storage device;
An image inspection program to be executed by a computer, wherein the acquired pixel value is set in a direction perpendicular to a tangent to a virtual curve drawn by a moving average of the pixels. An image inspection program for determining the quality of an inspection object using an image obtained by photographing the inspection object,
setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line and storing it in a storage device;
acquiring a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion and storing it in a storage device;
An image inspection program to be executed by a computer, wherein the acquired pixel value is set in a direction perpendicular to a straight line connecting pixels at preset intervals among the pixels. An image inspection program for determining the quality of an inspection object using an image obtained by photographing the inspection object,
setting a required observation portion at a desired location of the inspection object in the image using a straight line and/or a curved line and storing it in a storage device;
acquiring a linear or rectangular pixel value of a predetermined length extending in a predetermined direction from each of the continuous pixels constituting the observation-required portion and storing it in a storage device;
An image inspection to be executed by a computer, characterized in that the acquired pixel value is set by translating the observation target portion by a predetermined distance and connecting corresponding pixels before and after the translation. program .

Images