Detection Of Surface Defects And Geometric Quantities Of Auto Parts Based On Machine Vision

In the environment of continuous development and progress of industrial technology,product quality and safety issues have attracted more and more attention.The car seat belt buckle with good surface quality and precise dimensions can ensure the stable operation of the car seat belt system and effectively protect the life safety of the driver.Due to the irregular size of the car seat belt buckle and the smooth and curved surface,the irregular shape makes the quality inspection of the seat belt buckle difficult.Conventional manual detection methods are affected by personal factors,and stability and accuracy are difficult to guarantee.A real-time,high-efficiency and high-accuracy detection scheme for car seat belt buckles has become an urgent need.At present,machine vision technology is a new technology that is gradually developed in the field of industrial non-destructive testing.The workpiece is sampled through the cooperation of the camera lens light source,and then the sampled photos are processed by the host computer to extract the defects that need to be detected.Machine vision technology is developing as an important way to control product quality.It can achieve accurate assessment of the nature,size,location and quantity of the safety belt buckle defects,as well as the accurate measurement of the size of the seat belt buckle,while avoiding contact with the product to reduce damage.So as to ensure that the quality of the final product does not appear to be problematic.This subject takes several automobile seat belt buckle products as the inspection objects,and studies the methods and devices for detecting surface defects and geometric quantities of automobile seat belt buckles based on machine vision technology.Set up a machine vision inspection system,sample and save workpieces,analyze different defect features and classify them,conduct experimental verifications on different filtering and segmentation methods,and design image processing algorithms for defect detection and geometric measurement based on machine vision.Perform classification detection and extraction.The experimental results show that the algorithm has a higher than 98% recognition rate of safety belt buckle scratches,pits,gaps and other defects,which improves the yield and safety of use,reduces the cost of manual inspection,and realizes the detection of surface defects and dimensions of car seat belt buckles in engineering applications detection.