Nondestructive testing method for surface defects of mechanical parts based on machine vision

Abstract

In order to solve the problems of low detection accuracy and efficiency in the traditional surface defect detection methods of mechanical parts, a non-destructive testing method for surface defects of mechanical parts based on machine vision is proposed. This method designs an image acquisition architecture based on machine vision, and sets and selects camera parameters and light sources. According to the above framework, the distortion correction model is constructed by using inverse matrix method to complete the distortion correction processing of mechanical part image. Finally, the visual perception feature extraction of mechanical part image is carried out to complete the nondestructive testing of mechanical part surface defects. The experimental results show that, compared with the traditional detection methods, the proposed detection method has higher detection accuracy and efficiency, and is more practical.