Steel picture defect detection method in industrial production based on self-supervised contrast characterization learning technology

Abstract

The invention relates to a steel picture defect detection method in industrial production based on a self-supervised contrast characterization learning technology. Compared with the prior art, the defect that steel picture defect detection needs to depend on a large amount of marked data is overcome. The method comprises the following steps: acquiring and preprocessing an industrial production steel picture; constructing an upstream steel surface defect representation learning model; training an upstream steel surface defect representation learning model; constructing a downstream steel surface defect detector; training a downstream steel surface defect detector; acquiring a picture of industrial production steel to be detected; and obtaining an industrial production steel picture defect detection result. According to the method, a large amount of marked data is not needed, and good representation of defect data is obtained by performing self-supervised representation learning on a large amount of unmarked data sets, so that steel picture defect detection in industrial production is realized by using a small amount of marked steel defect pictures.

Description

technical field [0001] The invention relates to the technical field of industrial image detection, in particular to a method for detecting defects in steel product images in industrial production based on self-supervised comparative representation learning technology. Background technique [0002] Defect detection is an important link in industrial production. Steel surface defect detection refers to the detection of scratches, cracks, foreign objects, corrosion, holes and other problems on the steel surface. Traditional detection methods relying on manpower have disadvantages such as low efficiency, and the detection results are easily affected by human subjective factors. With the advancement of science and technology, some nondestructive testing methods have been proposed and applied to various defect detection fields. In the field of steel defect detection, technologies such as magnetic particle testing, penetrant testing, eddy current testing, ultrasonic testing, and X...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to solve the defect detection of steel image defects in the prior art that needs to rely on a large number of label data, and provide a method for detecting steel image defects in industrial production based on self-supervised comparative representation learning technology to solve the above problems

Images