JP5821708B2 - Defect inspection apparatus and defect inspection method - Google Patents

Description

  The present invention relates to a defect inspection method and a defect inspection apparatus that model a human visual mechanism.

  Patent Document 1 describes a defect detection method that models an inspection mechanism of a skilled worker. In Patent Document 1, a moving image of an inspection target is acquired by moving an imaging viewpoint along an operator's line of sight, and a feature amount is calculated by performing image processing on the moving image, so that a defect is detected. The defect candidate part which has possibility of applicable is specified. A detailed inspection is performed on the defect candidate part to determine whether or not the defect is an actual defect.

  In Patent Document 2, a technique for inspecting a wide field of view at low resolution using dark field illumination and re-inspecting a small size foreign object that is difficult to pass or fail at a high resolution using bright field illumination for a narrow field of view. Is disclosed.

JP 2010-223932 A JP 2009-162563 A

  However, in Patent Document 2, since it is not known where a minute foreign matter exists, there is a possibility that an oversight may occur if the entire inspection object is not inspected. If the entire inspection object is inspected with high resolution, the processing time is enormous and the tact time of the manufacturing process is not in time. If the processing is simplified in accordance with the tact time, there is a problem that the defect recognition rate is lowered.

  The present invention has been made against the background of the above circumstances, and an object thereof is to provide a defect inspection apparatus and a defect inspection method capable of detecting a defect at high speed and with high sensitivity.

  The defect inspection apparatus according to the first aspect of the present invention includes an imaging unit that captures an image of an inspection target, and the image is divided into a plurality of regions, and the luminance values of the pixels included in each region are averaged and divided. An image generation unit that creates an image; and a defect candidate extraction unit that extracts a defect candidate region by comparing a luminance value of one of the plurality of regions with a luminance value of a region existing around the region; Is provided.

  The defect inspection apparatus according to a second aspect of the present invention further includes a defect detection unit that detects a defect by performing image processing on the defect candidate area in the above-described apparatus.

  The defect inspection apparatus according to a third aspect of the present invention is characterized in that, in the above-described apparatus, the image generation unit divides the image into a lattice shape.

  In the defect inspection apparatus according to a fourth aspect of the present invention, in the above apparatus, the image generation unit includes a first processed image obtained by dividing the image into a plurality of regions, the first processed image, a phase, and a direction. And a second processed image divided into a plurality of different regions, and the divided image is generated using the first processed image and the second processed image.

  In the defect inspection method according to the fifth aspect of the present invention, an image of an inspection object is taken, the image is divided into a plurality of regions, and a luminance value of pixels included in each region is averaged to create a divided image. Then, a defect candidate region is extracted by comparing the luminance value of one of the plurality of regions with the luminance value of a region existing around the region, and image processing is performed on the defect candidate region. To detect defects.

  A defect inspection method according to a sixth aspect of the present invention is characterized in that, in the above method, the image is divided into a lattice shape.

  A defect inspection method according to a seventh aspect of the present invention includes a first processed image obtained by dividing the image into a plurality of regions, and a second process obtained by dividing the image into a plurality of regions having phases and directions different from those of the first processed image. An image is generated, and the divided image is generated using the first processed image and the second processed image.

  According to the present invention, it is an object to provide a defect inspection apparatus and a defect inspection method capable of reducing a processing time and detecting a defect with high sensitivity.

It is a figure which shows the structure of the defect inspection apparatus which concerns on Embodiment 1. FIG. It is a figure explaining a part of structure of the defect inspection apparatus which concerns on Embodiment 1. FIG. FIG. 3 is a flowchart showing a defect inspection method according to the first embodiment. It is a figure which shows an example of the image obtained by the imaging part of the defect inspection apparatus shown in FIG. It is a figure which shows the example of the grating | lattice which divides | segments the image of FIG. 4A. It is a figure which shows an example of the lattice division | segmentation image produced | generated by dividing | segmenting into the lattice shown to FIG. 4B. It is a figure which shows the defect candidate area | region detected based on the lattice division image of FIG. 4C. It is a figure which shows an example of the image obtained by the imaging part of the defect inspection apparatus which concerns on Embodiment 1. FIG. It is a figure which shows the example of the grating | lattice which divides | segments the image of FIG. 5A. It is a figure which shows an example of the lattice division | segmentation image produced | generated by dividing | segmenting into the lattice shown to FIG. 5B. It is a figure which shows the defect candidate area | region detected based on the lattice division | segmentation image of FIG. 5C. FIG. 10 is a flowchart showing a defect inspection method according to the second embodiment. It is a figure which shows an example of the image obtained by the imaging part of the defect inspection apparatus shown in FIG. It is a figure which shows the defect candidate area | region detected based on the divided image. It is a figure which shows an example of the image obtained by the imaging part of the defect inspection apparatus shown in FIG. It is a figure which shows the defect candidate area | region detected based on the divided image.

There is a theory that when detecting defects visually, humans are performing in two steps: discovery of defect candidate areas and inspection of the location. That is, when visual inspection is modeled from the human visual mechanism,
(1) Extraction of defect candidate areas based on global recognition of parts that are different from normal using peripheral vision. (2) It is considered to be two steps of judgment by examining defect candidate areas.

  The present invention models the extraction of defect candidate areas in the first step of these two steps. In the present invention, the image of the inspection object photographed by the camera is divided into a plurality of areas, and the luminance value of each area is averaged to generate a divided image with a reduced resolution. This is a new method that models the mechanism by which humans unconsciously recognize the difference from the normal state using these divided images. The determination of the presence or absence of a defect by examining the defect candidate area in the second step can be performed by a conventional method such as normal image processing or database collation.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The following description explains the embodiment of the present invention, and the present invention is not limited to the following embodiment. For clarity of explanation, the following description is omitted and simplified as appropriate. Moreover, those skilled in the art can easily change, add, and convert each element of the following embodiments within the scope of the present invention.

Embodiment 1 FIG.
A defect inspection apparatus according to Embodiment 1 of the present invention will be described with reference to FIGS. FIG. 1 is a diagram showing the configuration of the defect inspection apparatus according to the first embodiment. As shown in FIG. 1, the defect inspection apparatus 10 includes an imaging unit 11, a PC 12, and an illumination unit 13. FIG. 2 is a schematic diagram showing the configuration of the PC 12. As shown in FIG. 2, the PC 12 includes an image generation unit 14, a defect candidate extraction unit 15, and a defect detection unit 16.

  The imaging unit 11 captures an inspection object and generates an image. As the imaging unit 11, a camera such as a CCD can be used. When photographing with the imaging unit 11, it is possible to illuminate the inspection object with the illumination unit 13 as necessary.

  An image taken by the imaging unit 11 is taken into the PC 12. The visual mechanism by which humans discover defect candidate areas uses peripheral vision to recognize differences from the normal state. Peripheral vision is suitable for capturing movement and position by using the peripheral part of the human retina to look vaguely in the peripheral part of the visual field.

  In other words, the use of peripheral vision is not recognized using high-definition images. For this reason, in the present invention, the image captured by the image generation unit 14 is divided into a plurality of regions to generate a low-resolution divided image. This division imaging makes it possible to apply a human defect detection mechanism to a defect detection apparatus.

  The image generation unit 14 generates a divided image from the image captured by the imaging unit 11. First, an input image is divided into a plurality of regions. And the luminance value of the pixel contained in each area | region is averaged, and it is set as the luminance value of the said area | region. Then, the divided images are generated by combining the regions.

  The defect candidate extraction unit 15 extracts defect candidate areas using the divided images. A defect candidate area refers to an area that may contain a defect. As a specific process, the brightness value of one area of a plurality of areas is compared with the brightness value of an area existing around the area. A region that is smaller or larger than a predetermined value compared with the peripheral luminance value is set as a defect candidate region in which a defect exists in the vicinity of the region. The defect detection unit 16 examines the area determined as the defect candidate area using a conventional method such as image processing, and determines whether or not a defect exists.

  Here, the defect inspection method according to Embodiment 1 will be described with reference to FIGS. 3 and 4A to 4D. FIG. 3 is a flowchart for explaining the defect inspection method according to the first embodiment. As shown in FIG. 3, first, an imaging object 11 is imaged by the imaging unit 11, and the image is captured by the PC 12 (S11). FIG. 4A shows an example of an image obtained by the imaging unit 11. In the example shown in FIG. 4A, a plurality of circles are seen in the image, and a black line is present as a defect at the top of the circle at a position from the center right of the image.

  Next, the image captured by the PC 12 is divided into a plurality of areas (S12). FIG. 4B shows an example (broken line) of a grid for dividing the image of FIG. 4A. Here, an example in which an image is divided into a lattice shape will be described, but the shape of the divided region is not limited to the lattice. The size of the area to be divided can be arbitrarily determined from the size of the defect to be detected, the size of the visual field acquired by the imaging unit 11, and the like.

  Thereafter, the luminance values of the pixels included in each of the divided areas are averaged, and the divided areas are generated by combining the areas (S13). FIG. 4C shows an example of a divided image generated by dividing the grid shown in FIG. 4B. As shown in FIG. 4C, the luminance is averaged in each of the divided areas.

  Then, the brightness value in each area is compared with the brightness values of the surrounding areas (S14). A region whose luminance value in the region is greater than or equal to a predetermined value and larger or smaller than the surrounding luminance value is extracted as a defect candidate region (S15). FIG. 4D is a diagram showing defect candidate areas detected based on the lattice division image of FIG. 4C. In the example shown in FIG. 4D, a circle having a black line on the upper part of the circle shown in FIG. 4A is detected as a defect candidate region. Finally, a defect is detected by inspecting the inside of the defect candidate area in detail using image processing or the like as in the prior art (S16).

  With reference to FIG. 5A-5D, the example which used the pattern different from FIG. 4A-4D as a test object is demonstrated. FIG. 5A shows another example of an image obtained by the imaging unit 11. In FIG. 5A, a checker pattern is shown. As shown by a broken line in FIG. 5B, the image of FIG. 5A is divided into a grid pattern. Then, similarly to the above, the luminance values of the divided areas are averaged, and the divided images shown in FIG. 5C are generated by combining the areas. By comparing the brightness value of each area with the brightness value of the surrounding area, a defect candidate area is extracted as shown in FIG. 5D.

  As described above, the image obtained by photographing the inspection object is divided, and each region is averaged to use a divided image with a reduced resolution. Candidate regions can be extracted with high sensitivity. Since defects can be detected by examining only the defect candidate areas, it is possible to inspect at high speed.

Embodiment 2. FIG.
An inspection method according to Embodiment 2 of the present invention will be described with reference to FIG. FIG. 6 is a flowchart for explaining the inspection method according to the second embodiment. The second embodiment is different from the first embodiment in that divided images are obtained by integrating a plurality of processed images generated by changing the phase and direction of a lattice that divides an image of an inspection object into a plurality of regions. Is a point that generates

  Here, the phase of the grating means, for example, an angle at which the grating is rotated around the center point of the grating that divides the image of the inspection object into a plurality of regions. The direction of the lattice refers to the direction in which the columns or rows of the matrix lattice extend.

  As shown in FIG. 6, first, the imaging object 11 is photographed by the imaging unit 11 and the image is captured by the PC 12 (S21). Next, the image captured by the PC 12 is divided into a plurality of regions by the lattice (S22). Thereafter, the luminance values of the pixels included in each of the divided areas are averaged, and the respective processed areas are synthesized to generate a first processed image (S23).

  Then, the phase and direction of the grating are changed (S24). The degree to which the phase and direction of the grating are changed can be arbitrarily determined from the defect size and the like. Thereafter, the image acquired in S21 is divided into a plurality of regions with the changed phase and direction (S25), the luminance values of the pixels included in each divided region are averaged, and the respective regions are synthesized. A second processed image is generated (S26).

  Then, the first processed image and the second processed image are integrated to generate a divided image (S27). The luminance values in the divided images are compared (S28), and an area smaller than or larger than the luminance value of the surrounding area is extracted as a defect candidate area (S29). Finally, a defect is detected by inspecting the inside of the defect candidate area in detail using image processing or the like as in the prior art (S30).

  In the example illustrated in FIG. 6, the divided images are generated by integrating the two processed images of the first processed image and the second processed image. However, the present invention is not limited to this. A divided image may be generated by generating more than two processed images and integrating the processed images.

  FIG. 7A shows an example of an image obtained by the imaging unit 11 of the defect inspection apparatus 10 shown in FIG. FIG. 7B shows a divided image obtained by integrating a plurality of processed images whose phases and directions are changed from the image shown in FIG. 7A. In the example shown in FIG. 7B, the brightness value of the area from the center right of the divided image is higher than that of the other areas, and this area is detected as a defect candidate area.

  FIG. 8A shows another example of an image obtained by the imaging unit 11 of the defect inspection apparatus 10 shown in FIG. FIG. 8B shows a divided image obtained by integrating a plurality of processed images whose phases and directions are changed from the image shown in FIG. 8A. In the example shown in FIG. 8B, the lower left area of the divided image has a higher luminance value than other areas, and this area is detected as a defect candidate area. As described above, in any of the examples of FIGS. 7B and 8B, it is possible to detect a defect with high sensitivity.

  Note that the present invention is not limited to the above-described embodiment, and can be modified as appropriate without departing from the spirit of the present invention. For example, after determining a range in which luminance values are averaged using a lattice, the luminance values may be averaged after changing the phase and direction by rotating the image in each lattice. In this case, as in the first embodiment, the defect candidate region can be extracted by comparing the luminance of the lattice whose phase and direction are changed with the luminance of the surrounding lattice.

DESCRIPTION OF SYMBOLS 10 Defect inspection apparatus 11 Imaging part 12 PC
DESCRIPTION OF SYMBOLS 13 Illumination part 14 Image generation part 15 Defect candidate extraction part 16 Defect detection part

Claims (5)

An imaging unit for taking an image of the inspection object;
A first processed image obtained by dividing the image into a plurality of regions and averaging luminance values of pixels included in each region; and the image is divided into a plurality of regions having phases and directions different from those of the first processed image; An image generation unit that generates a second processed image obtained by averaging luminance values of pixels included in the region, and generates a divided image using the first processed image and the second processed image;
A defect candidate extraction unit that extracts a defect candidate area by comparing a luminance value of one area of the plurality of areas of the divided image with a luminance value of an area existing around the area;
A defect detection apparatus comprising:   The defect detection apparatus according to claim 1, further comprising a defect detection unit that detects a defect by performing image processing on the defect candidate area.   The defect detection apparatus according to claim 1, wherein the image generation unit divides the image into a grid pattern. Take an image of the inspection object,
A first processed image obtained by dividing the image into a plurality of regions and averaging luminance values of pixels included in each region; and the image is divided into a plurality of regions having phases and directions different from those of the first processed image; Generating a second processed image that averages the luminance values of the pixels included in the region;
Generating a divided image using the first processed image and the second processed image;
Comparing the brightness value of one area of the plurality of areas of the divided image with the brightness value of an area existing around the area, a defect candidate area is extracted,
Detecting defects by performing image processing on the defect candidate area;
Defect detection method. The defect detection method according to claim 4 , wherein the image is divided into a grid pattern.