Defect detection method and device based on cooling fin image recognition and medium

Abstract

The invention discloses a defect detection method and device based on cooling fin image recognition and a medium, and relates to the technical field of intelligent detection, and the method comprises the steps: collecting image data, carrying out the denoising, enhancement and geometric correction processing, and generating a preprocessed image; performing edge detection and contour feature extraction on the preprocessed image to generate structured contour data; carrying out geometric modeling on the structured contour data by adopting a triangulation algorithm to generate a three-dimensional geometric model; performing geometric deviation comparison on the three-dimensional geometric model, marking a geometric deviation area, and generating a defect marked three-dimensional geometric model; combining the defect marking three-dimensional geometric model with the preprocessed image, and performing surface texture comparison and defect prediction to generate a defect comprehensive result; and according to a defect comprehensive result, performing heat distribution anomaly analysis on the defect marking three-dimensional geometric model and the preprocessed image to generate a defect detection result. According to the invention, the effects of high precision, repeatability and quantifiability are achieved.

Images