Object surface flaw detection and analysis method and device based on photometric stereo

Abstract

The invention provides an object surface flaw detection and analysis method and device based on photometric stereo, and the method comprises the steps: obtaining pixel values of a to-be-detected object surface image of at least three different-angle light sources, and constructing a photometric stereo mathematical model; carrying out reflectivity calculation on the photometric stereo mathematical model, and constructing a reflectivity graph; calculating to obtain a gradient field of the surface of the to-be-measured object; performing Gaussian surface curvature calculation and average surface curvature calculation according to the gradient field of the surface of the to-be-measured object object to obtain a Gaussian surface curvature defect map and an average surface curvature defect map; and analyzing the reflectivity graph, the Gaussian surface curvature defect graph and the average surface curvature defect graph to obtain a detection result. The method avoids the problem that the traditional image processing technology is difficult to analyze the surfaces of various different objects to find out flaws, and has the characteristics of high detection accuracy and high universality.

Description

technical field [0001] The present application relates to the technical field of image detection, in particular to a method and device for detecting and analyzing object surface blemishes based on photometric stereo. Background technique [0002] Appearance inspection can find foreign matter, stains, blemishes, defects, etc., and prevent the outflow of defective products, but there is a limit to the accuracy of visual inspection. Full inspection not only consumes labor and cost, but also leads to accuracy deviation and human error due to individual differences. Moreover, it is difficult to find subtle flaws and stains. To maintain the quality, it is necessary to use tools such as microscopes to perform magnification inspections. When the number of points is small, microscopic inspection can be performed offline, but when thousands of points are to be inspected, a huge amount of labor is required, resulting in a significant drop in production efficiency. To balance quality ...

AI Technical Summary

Problems solved by technology

The surface material, texture, and illumination of the object to be tested are quite different. The design of traditional algorithms is difficult, the detection accuracy is low, and the universality is poor. For example, scratches and dents on the surface of lithium batteries are not easy to be detected; Deep learning technology has high requirements on hardware and requires a large number of rich samples

Images