Low-temperature butterfly valve surface defect detection method and system based on machine vision

The invention provides a low-temperature butterfly valve surface defect detection method and system based on machine vision, and belongs to the field of defect detection. The problem of low butterfly valve detection efficiency is solved. The method specifically comprises the following steps: S1, obtaining an original image, and obtaining a to-be-analyzed image; s2, backing up the picture to be analyzed; processing the to-be-analyzed data to obtain a primary analysis report; s3, acquiring an abnormal region set; traversing pixel points of the to-be-analyzed image, and extracting a surface anomaly set; judging whether the butterfly valve is available or not according to the abnormal area set and the surface abnormal set; s4, summarizing the primary analysis report and the secondary analysis report, and feeding back to the user; the image of the butterfly valve is acquired and processed, whether the butterfly valve is available or not is analyzed by combining the internal abrasion state of the butterfly valve, a