Metal multi-surface defect detection method under dynamic uncertain manufacturing scene

• Main Authors: JIANG YUHANG, YAN PING, SU YINGTAO, LIN JUNYAO, WEI YUNYAO, CHEN JIAN, ZHU KUANJUN
• Format: Patent
• Language: Chinese; English
• Published: 19.07.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a metal multi-surface defect detection method in a dynamic uncertain manufacturing scene, and the method comprises the steps: inputting an obtained metal surface image of a to-be-detected sample into a trained metal surface defect detection model, and obtaining a defect detection result of the metal surface image of the to-be-detected sample; according to the trained metal surface defect detection model, a reconstructed defect image and a real defect sample image are used as a defect sample training set through a training generation model, the defect sample training set is used for training the metal surface defect detection model, and therefore defect detection of the metal surface image of the sample to be detected is achieved. According to the method provided by the invention, the problems of missing detection and error detection caused by uncertain manufacturing scene and insufficient training samples in the existing defect detection method are effectively solved, and the generaliz