| The purpose of double row wire harness connector(hereinafter referred to as double row wire harness)detection is to detect the assembly unqualified defective wire harness assembly products,which are usually shown as wire-missing,wrong arrangement and reverse assembly.At present,detection in factory is mainly accomplished by workers,which has the following disadvantages: easy fatigue,high labor cost and low efficiency,thus it cannot meet the requirements of modern production.Based on machine vision inspection technology,this paper designs a set of visual inspection equipment for double harness connector so as to fulfill the detection of defective products.The main research contents of this paper are as follows:Firstly,according to the detection principle of machine vision and the detection requirements of double-row wire harness,this paper gives a detailed discussion and analysis to the selection of camera,lens,light source,support and light distribution mode,which belong to the hardware part.In order to solve the problem of distinguishing the front and back of double row wire harnesses and detecting the front-and-back same color wire-missing that cannot be solved by the common image acquisition device,a three-camera image acquisition device is designed,in which two cameras are responsible to collect the color sequence images of wire harnesses on the left and right respectively,and one camera to collect the terminal images of wire harnesses on the top.As to the problem of variety kinds of wire harnesses,this paper designs a bracing structure which can be adjusted from front to back and left to right.In the software part,the machine vision development environment is constructed for subsequent image processing algorithms,including the preparation of image acquisition window and human-computer interaction interface.Secondly,adaptive to ROI extraction algorithm,an algorithm of color sequence recognition is designed to obtain the best detection region and extract the image.In order to eliminate the image noise without destroying the contour information of the image,median filtering and bilateral filtering fusion are used to filter the obtained image.The bilinear interpolation algorithm is used to stretch the image and eliminatethe influence of the longitudinal inclination of the harness on the image segmentation.Transform RGB color space into HSL color space and increase S component multiple to enhance the saturation of the image,so as to solve the problem of misjudging similar color wires and enhance the contour features of the wire harness at the same time.Canny operator is used to detect the edge of the image and segment the partially overlapping and occluded wire harness through the calibrated wire width data.In order to eliminate the errors that may exist after the wire harness segmentation,a color extraction algorithm based on three-value filtering and color histogram is designed to improve the robustness of the final color extraction.Thirdly,due to the color sequence recognition algorithm cannot recognize the front and back of the terminals and the situation of front-and-back same color wire-missing,a recognition algorithm is designed according to the image features of the terminals.This method uses the idea of differential image hash algorithm to realize the function of distinguishing the front and back of the harness.Considering the different size of terminals,an image morphological processing algorithm is designed to automatically determine the size of the convolution kernel according to the length and width of the calibrated terminals,so as to detect the wire-missing of different types of terminals,and trigger the detection decision with template matching algorithm based on the square deviation.Finally,double-row wire harnesses of different specifications are used to verify the accuracy of the equipment.The detection rate of the defective products of the equipment reaches 100%,with the false alarm rate of common wire harnesses under10%,and the same color wire-missing under 15%. |
PDF Full Text Request