Research On Surface Defect Detection Of Automobile Stamping Parts Based On Machine Vision

Stamping parts are extensively used in the automobile manufacturing industry.And the methods for detecting the surface quality of many kinds of stamping parts include visual inspection,touch detection and oil detection.The shortcomings of artificial detection(such as the resolution of the human eyes,visual fatigue,etc.)make the enterprise cannot guarantee a high and stable pass rate and detection efficiency though the surface quality inspection of the work piece has been put into a large cost.Therefore,it is necessary to study the automatic detection technology for the surface defects of automobile stamping parts in time to improve the detection efficiency and quality.As a non-contact automatic inspection technology,machine vision inspection technology is widely used in recent years.It major principle is to imitate the human visual perception,obtain the initial image,save processing and judgment,so that the machine can have a visual function like human.Thus the machine can achieve detection,judgment,recognition,measurement and other functions.This method is an important way of detection and analysis of surface defects on the work piece.Therefore,the application of machine vision technology to the surface defect detection of stamping parts can improve the testing accuracy and automation level of the surface quality of stamping parts.The research content of this paper is the surface defect detection of stamping parts based on machine vision.And the detection object is the defect on the parts of automobile stamping parts.The defects mainly include the appearance defects: indentation and scratches,functional defects: small holes and marginal material.On the basis of careful analysis of the testing requirements,this paper mainly completes the following tasks:1.The research status and development trend of machine vision inspection technology at home and abroad are introduced;2.This paper analyzes the types and requirements of the surface defects of various automobile stamping parts,and designs the machine vision system for inspecting the surface defects of stamping parts.By using different lighting methods,the experimental and analytical comparison of the defects of the stamped parts is carried out,and the method of lighting for different types of defects is given.By analyzing the relationship between detection accuracy and camera resolution,an industrial camera with suitable resolution is selected.3.A digital image processing algorithm for targeted surface defects is proposed,which can effectively process the detected images and identify defects,so as to give the result information by analyzing the initial image features of different detected parts.The experimental software is programmed,and the validity of the defect detection algorithm is verified by processing the images collected on the experimental platform through the combined programming of HALCON and VS 2010.