DEFECT ANALYSIS DEVICE, DEFECT ANALYSIS METHOD, AND PROGRAM

WIPO Patent Application WO/2024/014102

A1

The present invention analyzes the classification of a defect on the basis of an image having captured the defect. This defect analysis device comprises: a model storage unit for storing a feature extraction model that is generated by performing learning on the basis of learning data obtained by adding information indicating classifications of flaws to defect images having captured therein flaws generated on the surface of an object, and that extracts, from an input image, an image feature quantity with which the distance between similar images becomes small; a representative point determination unit for determining representative points for the respective classifications on the basis of the image feature quantity extracted from the defect image by using the feature extraction model; and a relationship visualization unit for outputting information indicating the relationship between the classifications on the basis of the distances between the representative points.

OKANO YU (JP)
TAKADA SHINTARO (JP)
AKATSUKA TAKUMI (JP)
YOKOYAMA TETSUYA (JP)
TAKEMOTO SHIMPEI (JP)
OKUNO YOSHISHIGE (JP)

PCT/JP2023/017967

January 18, 2024

May 12, 2023

Click for automatic bibliography generation  

RESONAC CORP (JP)

G06T7/00; G06V10/762

WO2022130762A1	2022-06-23
WO2021199706A1	2021-10-07
WO2020004101A1	2020-01-02

JP2009068946A	2009-04-02
JP2000057349A	2000-02-25

OSHIRO YOSHITOMO, ENDO SATOSHI: "Improving accuracy of Fine-grained Classification using Contrastive Learning", PROCEEDINGS OF 20TH FORUM ON INFORMATION TECHNOLOGY (FIT2021), PART 3, 12 August 2021 (2021-08-12), pages 7 - 10, XP093128496

ITOH, Tadashige et al. (JP)

View/Download PDF      PDF Help

Previous Patent: INJECTION MOLDED BODY AND METHOD FOR PRODUCING SAME, METHOD FOR PRODUCING COMPOSITE FIBER, CNT-ATTAC...

Next Patent: METHOD AND DEVICE FOR MEASURING OPTICAL CHARACTERISTICS OF MULTI-CORE OPTICAL FIBER