Design Of The Wire Harness Terminal Crimping Appearance Inspection System Based On Machine Vision

With the development of China’s industrialization,the demand for wire harness is increasing,it is generally faced with the problem of technical backwardness in quality inspection,and the quality of the wire harness terminal crimping determines the quality of the wire harness.The existing crimping quality inspection methods are mainly divided into online crimping waveform detection and off-line cross-sectional image sampling,which have unclear inspection results and low efficiency,high miss rate And the quality of the crimping can be assessed directly by the crimped appearance.So,this paper designs a detection system of wire harness terminal crimping appearance based on machine vision.The main contents are as follows（1）According to the inspection standards of the wire harness terminal and the actual production requirements.Firstly,The common appearance defects of wire harness terminals are analyzed and the detection content are clarified Secondly,the hardware design of the visual acquisition system are carried out,including the camera,lens and light source.Finally,getting the good quality images,the image database of wire harness terminal is constructed through image difference technique,which provide image data support for detection algorithm design and verification（2）An image segmentation algorithm of the wire harness terminal is designed.For the general image segmentation algorithms take a long time and many interferences in the results This paper designs a segmentation algorithm based on the positioning reference translation.By extracting the positioning reference of the wire harness terminal obtains the target parts,with the help of the wire harness terminal’s production size.The segmentation experiments of wire harness terminals in different size and shape are carried out.The results show that the segmentation algorithm designed in this paper can complete the image segmentation of each part which need to be inspected,it means the segmentation algorithm is useful（3）The defect detection algorithms are designed.According to the different defect characteristics of each part,this paper designs A and C part defect detection algorithm based on the quantity and length of wire,B part defect detection algorithm based on contour extraction,and D part defect detection algorithm based on area feature.Finally,the four-part defect detection experiments are carried out in the offline image libraries P1000₀01,P1000₀02 and P1000₀03.The results show that the average accuracy rate is 99.80%and there is no missing detection,the average detection time of a single terminal is 10ms.These mean the algorithm has high accuracy and efficiency（4）An inspection software system is developed,it integrates the vision hardware with the image inspection algorithm,and performs dynamic detection experiment.Through the crimp test of 3600 HPC-001 wire harness terminals,the accuracy rate is 99.72%,the false positiveness rate is 0.28%,no missing-inspection.A single terminal takes an average of 80ms,which meets the design requirements.By comparing with the offline detection results,the reasons for the decrease in accuracy are analyzed.Through the improvement of the equipments’ performance.the pass rate increased from 98.97%to 99.89%,increased by 0.92%,which verifies the correctness of the analysis.Combining with the defective terminal images,providing the feasible measures for better performance of wire harness.