| CPC G06T 7/001 (2013.01) [G06F 18/2132 (2023.01); G06T 7/11 (2017.01); G06T 7/174 (2017.01); G06V 10/751 (2022.01); C23F 1/00 (2013.01); G06T 2207/30136 (2013.01)] | 10 Claims |
|
1. A machine vision-based automatic identification and rating method for low-magnification acid etching defect, used for automatically recognizing and ranking defects of a low-magnification acid-etched steel or steel billet or continuous casting billet sample after acid etching, comprising:
acquiring images of the low-magnification acid-etched sample according to a first pre-set condition to obtain a first image;
performing automatic image processing on the first image to obtain a second image;
performing image segmentation on the second image according to a second pre-set condition to obtain a third image, the image segmentation comprising taking the center of the second image as the center of a circle, performing concentric circle segmentation on the second image, and simultaneously performing quadrant segmentation on the second image according to a quadrant segmentation line of a pre-set quadrant to obtain the third image;
performing defect pattern recognition on the third image according to a pre-known defect type, and obtaining distribution data of defect patterns in the low-magnification acid-etched sample;
obtaining quantitative data of defect patterns in the low-magnification acid-etched sample according to the third image and the distribution data of defect patterns in the low-magnification acid-etched sample; and
ranking the defects in the low-magnification acid-etched sample according to the quantitative data of defect patterns in the low-magnification acid-etched sample.
|