US 12,243,216 B2
System and method for AI visual inspection
Martin Bufi, Waterloo (CA); and Raef Shehata, Waterloo (CA)
Assigned to Musashi Auto Parts Canada Inc., Waterloo (CA)
Appl. No. 17/765,222
Filed by Musashi Auto Parts Canada Inc., Arthur (CA)
PCT Filed Sep. 30, 2020, PCT No. PCT/CA2020/051306
§ 371(c)(1), (2) Date Mar. 30, 2022,
PCT Pub. No. WO2021/062536, PCT Pub. Date Apr. 8, 2021.
Claims priority of provisional application 62/907,767, filed on Sep. 30, 2019.
Prior Publication US 2022/0366558 A1, Nov. 17, 2022
Int. Cl. G06T 7/00 (2017.01); B25J 19/02 (2006.01); G01N 35/00 (2006.01); G05B 19/418 (2006.01); G06T 3/4038 (2024.01); G06V 10/764 (2022.01); G06V 10/82 (2022.01); H04N 23/695 (2023.01)
CPC G06T 7/0004 (2013.01) [B25J 19/023 (2013.01); G01N 35/0099 (2013.01); G05B 19/41875 (2013.01); G06T 3/4038 (2013.01); G06V 10/764 (2022.01); G06V 10/82 (2022.01); H04N 23/695 (2023.01); G05B 2219/32368 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06V 2201/06 (2022.01)] 19 Claims
OG exemplary drawing
 
1. A system for visual inspection of an article, the system comprising:
a camera for acquiring image data of an article under inspection;
a node computing device for receiving the image data from the camera and analyzing the image data using a defect detection model trained to detect at least one defect type, the defect detection model comprising a machine-learning based object detection model configured to receive the image data as an input and generate defect data describing a detected defect as an output; and
a programmable logic controller (“PLC”) device for receiving the defect data from the node computing device;
wherein the defect data includes a defect class corresponding to a defect type, a defect location, and a confidence level, and wherein the defect detection model is configured to perform multiclass classification for classifying instances into one of three or more classes, wherein the classes include at least two defect types.