Quality Inspection of Workpiece Camouflage Spraying Based on Improved YOLOv3-tiny

Existing workpiece defect detection is mostly aimed at surface defects, or car surface defects with a single background color, and there is no professional quality inspection for camouflage spraying. Camouflage spraying is very common in industrial fields, such as large industrial structural parts, car boxes, etc. The quality inspection of camouflage spraying has a complex background and small defect targets. It has higher requirements for surface inspection technology and more difficult feature extraction. Since there is no open source data set related to camouflage spraying, we combined the actual industrial control environment needs to build a camouflage spraying data set through web crawlers, self-made methods, etc., increased the amount of data through data enhancement, and combined with computer vision, a method for detecting the quality of camouflage spraying on workpieces with improved YOLOv3-tiny algorithm is proposed to realize the detection of camouflage spraying quality on the surface of workpieces, which is of great significance to the maintenance of actual industrial control components.