Research On PCB Bare Board Defect Detection Technology Based On Machine Vision

In recent years,with the continuous development of the level of modern information technology,people put forward higher requirements for the quality of the electronic products.Higher product quality requirements need to update equipment.As the carrier of electronic products,the quality detection of printed circuit board(PCB)has become a key problem in the electronic manufacturing industry.The traditional PCB defect detection is mainly through the manual detection or functional detection to achieve,but this method has many disadvantages.In order to improve the efficiency and accuracy of defect detection,the paper designs a PCB defect detection system based on machine vision,and realizes the high speed and high precision detection of PCB bare board defects.Above all,the paper introduces the background and significance of the research,summarizes the method of PCB defect detection and research status at home and abroad.Then,according to the requirement of PCB defect detection,the overall design scheme of PCB defect visual inspection system is put forward.PCB image acquisition module,platform motion control module and image processing module are described in detail.PCB image preprocessing is an important step in defect visual inspection.In the paper,the PCB color image is transformed into PCB gray image.In order to make the PCB image and the standard PCB image to be accurately registered,the perspective transformation algorithm is used to rotate the PCB gray image.As the difference between the PCB pad and the wire gray is not obvious,the contrast enhancement algorithm is used to enhance the PCB image.Considering the noise composition of the industrial field PCB image,a hybrid denoising algorithm based on wavelet transform and adaptive mean filtering is proposed.The experiment shows that the algorithm can effectively remove the PCB image noise and keep the edge information of the image well.Based on the gray difference of the PCB pad background and the wire target,a threshold segmentation algorithm based on wavelet decomposition and gray histogram is studied to accurately segment the PCB image pad and wire area.Aiming at the possible defects of PCB board,a visual inspection and recognition algorithm for PCB bare board defect is designed and implemented.In the paper,the PCB bare board defects mainly include short circuit,open circuit,notch,burr,hole,and so on.The algorithm first corrode the target area of the bare PCB image,and then transform the corrugated image to obtain the distance gray image.The image is compared with the standard PCB distance gray image to detect the defect area.Then,by extracting the contours,the number of Euler numbers,the connected domain and the area of the defect area,the classification decision tree is constructed to identify the defect type.The experimental results show that the algorithm can effectively detect and identify PCB defects,and has high accuracy and good robustness.Finally,the software platform of PCB bare board defect visual inspection system is designed and developed.The design and development of each function module of the system are described in detail.The problem in the actual debugging is solved.Ensure the system can run fast and accurately,and realize the real-time online detection of PCB bare board defect.