Research And Development Of Capacitor Appearance Defect Detection System Based On Machine Vision

Now,with the overall improvement of industrial automation,the market demand for electronic components has become larger and larger.Capacitors are one of the most commonly used electronic components.Major manufacturers have already realized the scale and mass production of capacitors in the face of their increasing market demand.However,for the appearance defects of capacitors,most manufacturers still use manual visual inspection,which seriously affects the production efficiency and sales of capacitors.In this paper,according to the production standards and requirements,a prototype of a real-time detection system for capacitor appearance defects is designed according to the appearance defects generated in the CCB capacitor production process.Firstly,the background and significance of the research are briefly introduced,and the shortcomings of artificial naked eye detection are analyzed.At the same time,it gives an overview of machine vision technology,and summarizes the research status of appearance defect detection at home and abroad and the development trend of appearance defect detection.Secondly,it analyzes the appearance defects generated in the capacitor production process,outlines the typical machine vision inspection system,and gives a detailed introduction to the structure of the machine vision-based capacitor appearance defect detection system studied in this paper.The image processing software used in this topic was introduced to select the camera lens and design the correct light source system.According to the actual engineering conditions,the design and construction of the mechanical platform is completed from the perspective of facilitating the detection of capacitor appearance defects.Then,according to the detection system flow,image acquisition,image filtering,image enhancement,morphology processing,matching positioning,edge detection,Blob analysis,etc.are introduced in turn.In the image enhancement,matching positioning and edge detection,the algorithm is improved and designed to improve the detection accuracy and real-time.In terms of image enhancement,the best histogram equalization algorithm is selected to improve the gray level merging problem,which makes the capacitor appearance defect detection more accurate.In the aspect of matching and positioning,the image is directly sampled through analysis and comparison.Improve the real-time detection;in the aspect of edge detection,improve the traditional Canny operator,replace the Gaussian filter with fast median filtering,improve the amplitude calculation method,and increase the accuracy and speed of detection.Finally,a brief introduction to the human-computer interaction interface and prototype ofthe capacitor appearance defect detection system is given.Run the prototype to verify the image enhancement,matching positioning and edge detection in the detection system,and check the speed and accuracy of the prototype detection.The test results show that the detection system can accurately and quickly detect the appearance defects of the capacitor,and the improvement work can further improve the detection accuracy and real-time.