Self-similarity analysis for defect detection on patterned industrial objects

Patent metadata

Field	Value
Publication number	US-10726540-B2
Application number	US-201715785904-A
Country	US
Kind code	B2
Filing date	Oct 17, 2017
Priority date	Oct 17, 2017
Publication date	Jul 28, 2020
Grant date	Jul 28, 2020

Abstract

A method for object defect detection includes receiving digital data representing an image of an object with a repeated pattern. The method identifies a part of the image of the object as defined by a sample window of the digital data. The method generates one or more functions from at least the part of the image, wherein each of the one or more functions corresponds to one component of a pixel contained in the part of the image. Responsive to performing self-similarity analytics on the one or more functions, the method identifies a defect area of the object.

First claim

What is claimed is:

1. A computer-implemented method for object defect detection, comprising:
receiving, by one or more processors, digital data representing an image of an object with a repeated pattern;
identifying, by one or more processors, a part of the image of the object as defined by a sample window of the digital data, wherein the sample window of the digital data includes a portion of the repeated pattern;
generating, by one or more processors, one or more functions from the part of the image, wherein each of the one or more functions corresponds to one component of a pixel contained in the part of the image in a single axis direction across the repeated pattern vertical to a periodical direction;
performing, by one or more processors, self-similarity analytics on the one or more functions from the part of the image of the object to a plurality of functions from a plurality of parts of the image of the object as defined by a plurality of sample windows of the digital image; and
responsive to determining the one or more functions exceeds a minimum deviation from the plurality of functions, identifying, by one or more processors, the part of the image of the object as a defect area of the object.

2. The method according to claim 1 , wherein identifying a part of the image comprises:
determining, by one or more processors, an original sample window of digital data confining a part of the image;
extending, by one or more processors, the original sample window to generate at least one other sample windows, wherein each of the other sample windows covering the original sample window; and
determining, by one or more processors, the plurality of parts of the image, wherein each of the plurality of parts of the image being confined by the original sample window and each of the other sample windows respectively.

3. The method according to claim 1 , wherein identifying a part of the image comprises:
determining, by one or more processors, an original sample window of digital data confining a part of the image;
generating, by one or more processors, at least one other sample window based on a pre-defined interval length, wherein each of the other sample windows partially covering the original sample window; and
determining, by one or more processors, the plurality of parts of the image, wherein each of the plurality of parts of the image being confined by the original sample window and each of the other sample windows respectively.

4. The method according to claim 1 , wherein the generated one or more functions are related to location information of pixels contained in the determined part of the image and values of components of the pixels.

5. The method according to claim 1 , wherein generating one or more functions from the part of the image comprises:
extracting, by one or more processors, values of one or more components of pixels contained in the part of the image; and
generating, by one or more processors, functions from the at least a part of the image based on the extracted values, wherein each of the functions corresponding to one of the components.

6. A computer program product comprising:
one or more computer readable storage media and program instructions stored on at least one of the one or more storage media, the program instructions comprising:
program instructions to receive digital data representing an image of an object with a repeated pattern;
program instructions to identify a part of the image of the object as defined by a sample window of the digital data, wherein the sample window of the digital data includes a portion of the repeated pattern;
program instructions to generate one or more functions from the part of the image, wherein each of the one or more functions corresponds to one component of a pixel contained in the part of the image in a single axis direction across the repeated pattern vertical to a periodical direction;
program instructions to perform self-similarity analytics on the one or more functions from the part of the image of the object to a plurality of functions from a plurality of parts of the image of the object as defined by a plurality of sample windows of the digital image; and
program instructions to, responsive to determining the one or more functions exceeds a minimum deviation from the plurality of functions, identify the part of the image of the object as a defect area of the object.

7. The computer program product of claim 6 , wherein program instructions to identify a part of the image comprises program instructions, stored on the one or more computer readable storage media, which when executed by a processor, cause the processor to:
determine an original sample window of digital data confining a part of the image;
extend the original sample window to generate at least one other sample windows, wherein each of the other sample windows covering the original sample window; and
determine the plurality of parts of the image, wherein each of the plurality of parts of the image being confined by the original sample window and each of the other sample windows respectively.

8. The computer program product of claim 6 , wherein program instructions to identify a part of the image comprises program instructions, stored on the one or more computer readable storage media, which when executed by a processor, cause the processor to:
determine an original sample window of digital data confining a part of the image;
generate at least one other sample window based on a pre-defined interval length, wherein each of the other sample windows partially covering the original sample window and
determine the plurality of parts of the image, wherein each of the plurality of parts of the image being confined by the original sample window and each of the other sample windows respectively.

9. The computer program product of claim 6 , wherein the generated one or more functions are related to location information of pixels contained in the determined part of the image and values of components of the pixels.

10. The computer program product of claim 6 , wherein program instruction to generate one or more functions from the part of the image comprises program instructions, stored on the one or more computer readable storage media, which when executed by a processor, cause the processor to:
extract values of one or more components of pixels contained in the part of the image; and
generate functions from the at least a part of the image based on the extracted values, wherein each of the functions corresponding to one of the components.

11. A computer system comprising:
one or more computer processors;
one or more computer readable storage media; and
program instructions stored on the computer readable storage media for execution by at least one of the one or more computer processors, the program instructions comprising:
program instructions to receive digital data representing an image of an object with a repeated pattern;
program instructions to identify a part of the image of the object as defined by a sample window of the digital data, wherein the sample window of the digital data includes a portion of the repeated pattern;
program instructions to generate one or more functions from the part of the image, wherein each of the one or more functions corresponds to one component of a pixel contained in the part of the image in a single axis direction across the repeated pattern vertical to a periodical direction;
program instructions to perform self-similarity analytics on the one or more functions from the part of the image of the object to a plurality of functions from a plurality of parts of the image of the object as defined by a plural

Assignees

• IBM

Inventors

• Huang Jing Chang
• Sun Wei
• Yan jun chi
• Yao Ren Jie
• Zhu Jun

Classifications

• G01N21/8851 Primary

  Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges (G01N21/8806 and G01N21/93 - G01N21/95692 take precedence; optical measurement of dimensions G01B11/00; optical scanning G02B26/10; image transformation G06T3/00; computerised image enhancement G06T5/00; image processing per se for flaw detection G06T7/0002) · CPC title

• G06T7/001 Primary

  using an image reference approach · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G01N2021/8864

  Mapping zones of defects · CPC title

• G06T7/0004

  Industrial image inspection · CPC title