A pattern recognition and matching method and system for irregular defect pictures

Abstract

The present invention provides a method and system for pattern recognition and matching of irregular defect pictures, comprising: acquiring steel surface defect images in real time and selecting defect distribution; identifying and matching the defect distribution obtained in the step of selecting defect distribution; displaying the matching result and performing Correlation Analysis. The invention can improve the identification and analysis ability of defect distribution features, lay a solid foundation for further analysis and mining of defect images, and meet the needs of enterprises to improve product quality and intelligence. Moreover, the invention can be extended to other fields related to image processing, and has wide application prospects.

Description

technical field [0001] The invention relates to image recognition and matching technology, in particular, to a method and system for pattern recognition and matching of pictures with irregular defects on steel surfaces. Background technique [0002] Steel surface defects are caused by various reasons during the production process, and these defects may directly affect the performance and quality of the final product. Therefore, in recent years, users in related industries have become more and more stringent on the surface quality of steel, and the surface quality of steel has become a key indicator of competition in the steel industry. [0003] Most of the existing steel surface defect detection adopts manual visual detection method, which is difficult to ensure the real-time and accuracy requirements of the detection system. In recent years, with the rapid development of machine vision technology, surface defect detection systems based on machine vision have begun to repla...

AI Technical Summary

Problems solved by technology

[0004] However, the current steel surface defect detection still has the following deficiencies: the detection of surface defects is limited to the detection and classification of steel surface defects; the analysis of defect distribution characteristics is limited to the identification of distribution areas, such as figure 1 As shown; the lack of tracking and identification of the whole process of defects is not conducive to the analysis of the cause of defects; the defect feature library and template library are both built on the basis of single defects, and the analysis of multiple defects with specific distribution is insufficient, etc.

Images