US 12,236,576 B2
Workpiece surface defect detection device and detection method, workpiece surface inspection system, and program
Akira Yahashi, Kobe (JP); Yoshihito Souma, Sakai (JP); Taizo Wakimura, Kishiwada (JP); Ryuichi Yoshida, Sakai (JP); and Shota Ueki, Toyonaka (JP)
Assigned to Konica Minolta, Inc., Tokyo (JP)
Appl. No. 17/639,396
Filed by Konica Minolta, Inc., Tokyo (JP)
PCT Filed Oct. 2, 2019, PCT No. PCT/JP2019/038908
§ 371(c)(1), (2) Date Mar. 1, 2022,
PCT Pub. No. WO2021/064893, PCT Pub. Date Apr. 8, 2021.
Prior Publication US 2022/0335586 A1, Oct. 20, 2022
Int. Cl. G06T 7/00 (2017.01); G06T 7/11 (2017.01); G06T 7/70 (2017.01); G06T 11/00 (2006.01)
CPC G06T 7/001 (2013.01) [G06T 7/11 (2017.01); G06T 7/70 (2017.01); G06T 11/00 (2013.01); G06T 2207/10152 (2013.01); G06T 2207/20132 (2013.01); G06T 2207/30164 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A workpiece surface defect detection device comprising:
a first hardware processor that:
obtains a plurality of images indicating a portion to be measured of a workpiece serving as a target of detection of a surface defect, in a state where a bright-and-dark pattern of an illumination device is moved relative to the workpiece;
extracts a tentative defect candidate for each of the plurality of images that has been obtained;
determines the tentative defect candidate as a defect candidate, when among a plurality of images from which the tentative defect candidate has been extracted, a number of images including the tentative defect candidate is greater than or equal to a threshold that has been set in advance;
cuts out, for each of a plurality of the defect candidates determined, a predetermined region around the defect candidate as an estimated region from a plurality of images including the defect candidate, and generates a plurality of estimated region image groups;
combines a plurality of images including the defect candidate that has been determined, and generates a composite image by combining a plurality of estimated region images that has been generated, and generating the composite image, for each of the plurality of the defect candidates; and
detects a defect on a basis of the composite image that has been generated, wherein
when generating the composite image, the first hardware processor aligns the plurality of estimated region images to gain a maximum evaluation value from among a plurality of combinations obtained by respectively shifting the center coordinates of each of the plurality of images in X-coordinate and Y-coordinate directions.