Research On Surface Roughness Detection Methods Of Electrical Connector Shell Based On Machine Vision

Electrical connector assembly is an important component of the control and drive system in the field of aerospace, industry, precision machinery and even household appliances. Electric connector shells’ sealing property corrosion resistance, and even the quality of electrical connector assembly, however, is seriously impacted by its surface roughness. Taking the electric connector shells as the object, a method based on machine vision is proposed in this paper for detecting the surface roughness efficiency and accuracy. The proposed method has higher detection efficiency and accuracy compared with manual detection, and provides a basis for the surface roughness detection of similar part in the future.According to the course of image acquisition, feature extraction, establishment of relational model and experiment of surface roughness detection, in-depth theoretical research and experimental analysis are studied on the key technologies. The main research contents of this paper are as follows:1. Acquisition of high quality images. Taking the high quality image acquisition as the object, the light source and lighting system, parameters of machine vision are analyzed, and an electric connector shell surface defect detection based on single line array CCD is proposed.2. Texture defect detection. Surface defects of electrical connector shell greatly affect the extraction of surface roughness parameters. In this paper, it is proposed to detect the surface defects of the shell before the characteristic parameter extraction of surface roughness. The image is classified according to the target area and the background area of the defect detection. On the basis of the characteristics of the electric connector shells’ texture defects, this paper proposes a texture defect detection method based on wavelet analysis.3. characteristic parameter extraction of surface roughness. Two kinds of theories about the surface roughness are studied: gray level co-occurrence matrix and fractal theory, and so did the method of extracting characteristic parameters. The value of the three structural factors of the gray level co-occurrence matrix is analyzed in detail.4. Relational model establishment for surface roughness. 14 characteristic parameters which characterize the surface roughness are respectively carried out by the least square method. Optimal choice of goodness of fit parameter sum of variance to establish the model of the relationship. According to the model, the surface roughness of the shell is analyzed, and the experimental results show that the detection error of the surface roughness is less than 7%.