Research On Automobile Connector Pin Detection System Based On Machine Vision

Car switch is a key component of the car parts,used to control the car’s rearview mirror,door,child lock,etc.In the process of automobile switch detection,the wiring harness should be connected first,and the wiring harness is connected by the connector.If the connector pin deflection,bending,will affect the male and female ends of the connector,serious will cause pin break,connection is not reliable and other consequences.Therefore,it is necessary to detect the pins before wiring harness connection.The traditional manual detection method has high labor cost and low efficiency,which is not conducive to realizing automatic production.In this paper,a scheme of connector terminal detection based on machine vision is proposed.And the machine vision detection algorithm,the research experiment,the final realization of the terminal detection.Specific research contents are as follows:(1)Through the analysis of the connector structure and combined with the detection requirements,the overall detection scheme that firstly divides the connector contour as the boundary and then identifies the pin is determined.(2)The goal is to obtain high-quality pictures with obvious pin features,uniform illumination and less noise interference.I selected the camera,designed the lighting scheme,built the visual detection experiment platform,and completed the picture collection.(3)With the recognition and segmentation of connector outer contour as the goal,the segmentation method based on edge and the segmentation method based on template matching are experimented.Through the analysis of the experimental results,the appropriate segmentation method is determined.(4)With pin recognition as the goal,the pin recognition and detection scheme based on BLOB analysis is determined,and the key and difficult problems that need to be solved in the process of image processing are summarized,such as noise problem,pin feature loss,degradation or interference caused by pin oxidation,which is difficult to distinguish.To solve these problems,the image grayscale processing algorithm research,image denoising algorithm research,spot feature extraction algorithm research and so on.The advantage of corner detection algorithm in spot feature extraction is emphatically studied.A method combining image fusion and morphological processing was proposed to enhance the features of pins,which solved the problems of pin feature loss and the difficulty in distinguishing bright spots from stray spots of pins.Finally,through the Blob analysis method,the pin identification is completed.By calculating the offset of each pin,the pin bending detection is completed.To sum up,this paper completed the design of automobile connector detection system based on machine vision,the system can meet the detection needs,has a certain engineering practical value.