Defect inspection apparatus, defect inspection method, and non-transitory computer readable medium

1. A defect inspection apparatus, comprising:

a memory section, storing a learned model and an internal parameter set in the learned model;

an acquisition section, configured to acquire an image of an inspection object photographed under predetermined conditions;

a preprocessing section, generating a predetermined preprocessing filter according to a feature value in a preprocessing image being an image of the inspection object that is acquired by the acquisition section and that comprises a defect, and a feature value corresponding to the internal parameter, and generating a preprocessed image by applying the generated preprocessing filter to an inspection image being an image of the inspection object that is acquired by the acquisition section and converting the inspection image; and

an inspection section, inspecting the preprocessed image for presence or absence of the defect of the inspection object by using the stored learned model,

wherein the preprocessing section generates the preprocessing filter for converting the preprocessing image so that a width of the feature value of a defective region including the defect in the preprocessing image is comprised in a range of the feature value corresponding to the internal parameter.

2. The defect inspection apparatus according to claim 1 , wherein

the preprocessing section generates the preprocessing filter for converting the preprocessing image so that a width of a pattern of a background region other than the defective region in the preprocessing image deviates from the range of the feature value corresponding to the internal parameter.

3. The defect inspection apparatus according to claim 1 , wherein

the preprocessing section receives a designation of the defective region in the preprocessing image from outside and generates the preprocessing filter based on the received defective region.

4. The defect inspection apparatus according to claim 2 , wherein

the preprocessing section receives a designation of the defective region in the preprocessing image from outside and generates the preprocessing filter based on the received defective region.

5. The defect inspection apparatus according to claim 1 , wherein

the preprocessing section generates the preprocessing filter enlarging or reducing the preprocessing image based on the defective region.

6. The defect inspection apparatus according to claim 2 , wherein

the preprocessing section generates the preprocessing filter enlarging or reducing the preprocessing image based on the defective region.

7. The defect inspection apparatus according to claim 1 , wherein

the preprocessing section causes a portion of the preprocessing image that comprises at least the defective region to change according to a plurality of enlargement/reduction ratios and generates a feature detection image with respect to at least the portion;

the preprocessing section generates the preprocessing filter enlarging or reducing the preprocessing image according to the enlargement/reduction ratio when a difference between a density of the feature detection image within the defective region and a density of the feature detection image outside the defective region is the greatest when changing with the plurality of enlargement/reduction ratios;

wherein the feature detection image is displayed in variable density so that the density of a region indicating a feature corresponding to the internal parameter is larger as compared with other regions.

8. The defect inspection apparatus according to claim 2 , wherein

the preprocessing section causes a portion of the preprocessing image that comprises at least the defective region to change according to a plurality of enlargement/reduction ratios and generates a feature detection image with respect to at least the portion;

the preprocessing section generates the preprocessing filter enlarging or reducing the preprocessing image according to the enlargement/reduction ratio when a difference between a density of the feature detection image within the defective region and a density of the feature detection image outside the defective region is the greatest when changing with the plurality of enlargement/reduction ratios;

wherein the feature detection image is displayed in variable density so that the density of a region indicating a feature corresponding to the internal parameter is larger as compared with other regions.

9. The defect inspection apparatus according to claim 1 , wherein

the preprocessing section generates the preprocessing filter flattening brightness of the preprocessing image.

10. The defect inspection apparatus according to claim 9 , further comprising a display section displaying the preprocessed image.

11. The defect inspection apparatus according to claim 10 , wherein

the display section displays the preprocessed image and a detection result image from which the defect is extracted in a manner enabling comparison.

12. The defect inspection apparatus according to claim 1 , comprising:

an input section receiving a user selection of the preprocessing image used for generating the preprocessing filter from the image of the inspection object that is acquired by the acquisition section and that comprises the defect;

wherein the preprocessing section generates the predetermined preprocessing filter according to the feature value in the preprocessing image and the feature value corresponding to the internal parameter.

13. The defect inspection apparatus according to claim 1 , comprising:

a sequence input section receiving a user selection of which of a plurality of processing sequences is to be executed,

wherein the plurality of processing sequences comprise a preparation process sequence in which the preprocessing section generates the preprocessing filter and an operation process sequence in which the generated preprocessing filter is applied to the inspection image to inspect presence or absence of the defect comprised in the inspection object.

14. The defect inspection apparatus according to claim 13 , wherein the operation process sequence executes a process that updates the preprocessing filter by the preprocessing section according to content of an inspection result regarding whether or not the defect of the inspection object can be identified.

15. A defect inspection method, wherein a computer executes steps of:

storing a learned model and an internal parameter set in the learned model;

acquiring an image of an inspection object photographed under predetermined conditions;

generating a predetermined preprocessing filter according to a feature value in a preprocessing image being an image of the inspection object that is acquired in the step of acquiring and that comprises a defect, and a feature value corresponding to the internal parameter, and generating a preprocessed image by applying the generated preprocessing filter to an inspection image being an image of the inspection object that is acquired in the step of acquiring and converting the inspection image; and

inspecting the preprocessed image for presence or absence of the defect of the inspection object by using the stored learned model,

wherein the preprocessing filter is generated for converting the preprocessing image so that a width of the feature value of a defective region including the defect in the preprocessing image is comprised in a range of the feature value corresponding to the internal parameter.

16. A non-transitory computer readable medium that stores a program, wherein a computer executes the program to:

store a learned model and an internal parameter set in the learned model;

acquire an image of an inspection object photographed under predetermined conditions;

generate a predetermined preprocessing filter according to a feature value in a preprocessing image being an image of the inspection object that is acquired by the unit for acquiring and that comprises a defect, and a feature value corresponding to the internal parameter, and for generating a preprocessed image by applying the generated preprocessing filter to an inspection image being an image of the inspection object that is acquired by the unit for acquiring and converting the inspection image; and

inspect the preprocessed image for presence or absence of the defect of the inspection object by using the stored learned model,

wherein the preprocessing filter is generated for converting the preprocessing image so that a width of the feature value of a defective region including the defect in the preprocessing image is comprised in a range of the feature value corresponding to the internal parameter.