Abstract:Surface defect detection is one of the key links in the quality inspection of seat belt. At present,low efficiency and poor stability exist in artificial detection,so actual detection needs can not meet.The seat belt surface defect detection algorithm was explored based on image processing algorithms to meet the actual demand, and the seat belt surface defect detection system based on machine vision was constructed. Aiming at the characteristics of surface defect of seat belt, a method of feature extracting based on spectrum analysis was presented, and the selection of the filter and the calculation of the parameters were discussed. The experimental results show that the spectrum analysis method has better effect on the detection of surface defects than other detection methods, which can meet the requirements of realtime and accuracy.