Research On Automatic Detection System Of PCB Defects Based On Machine Vision

In recent years,with the deepening of research in machine vision technology in the industrial field,the related applications of machine vision have attracted more and more attention.PCB defect detection,as an important application of machine vision in the industrial field,is also applied to PCB production lines on a large scale.With the technological innovation and industrial upgrading,the requirements for PCB production process are getting higher and higher,but high-end AOI equipment is still monopolized by foreign companies.Its high price has forced many small-scale PCB manufacturers in China to detect their PCB products through manual testing.With the rapid development of many emerging technologies such as Industry 4.0 and Internet of Things,manufacturing The industry is bound to enter a new stage,and it is of great significance and value to study the realization of PCB vision detection system based on machine vision.Therefore,this topic studies to realize an automatic detection system for PCB defects based on machine vision.This paper first discusses the research background and significance of the subject,and summarizes the research status of PCB defect detection at home and abroad.Explain how the system works and design the overall architecture of the system.Choose the right hardware based on the major external factors that affect system performance.The preprocessing algorithm of PCB image is completed.Preprocessing is a very important part of PCB defect detection system.This paper first converts the color image acquired by CCD camera into grayscale image,and then enhances the image of grayscale image.Image smoothing and other processing,based on the characteristics of PCB image,based on the study of the classical Otsu algorithm,an improved algorithm combining genetic algorithm and Otsu is proposed to segment the PCB image,and the algorithm is effective and reliable..An algorithm for automatic detection and identification of PCB defects is proposed and implemented.The algorithm firstly registers the PCB image to be tested with the template by the improved SURF algorithm,so that they are aligned in space,and then using mathematical morphology to process the PCB image,and obtain the position of the defect through differential operation.The results proves the algorithm can detect commonly PCB defects,and has high accuracy.Finally,the system is built,the software platform of the PCB defect automatic detection system is designed and implemented,and the functions of each module in the system are expounded.The software interface is simple and clear,and the human-computer interaction performance is good.The detection and detection system has high recognition accuracy and reliable performance.