Research On Stencil Defects System Based On Machine Vision

With the rapid development of electronic technology,the complexity and integration of PCB board circuits is increasing,and surface mount technology(SMT)has become the most commonly used technology in electronic assembly process.The main function of the SMT stencil is to transfer the exact amount of solder paste to the exact position on the PCB.Therefore,the quality of the openings in the steel mesh will directly affect the subsequent process and the quality of the board soldering.The traditional steel mesh detection method is manual detection.Nowadays,the number and size of the opening of the steel mesh have caused great difficulties for manual inspection,and the quality and efficiency have been unable to meet the needs of the industry.Therefore,this thesis designs a set of SMT stencil detection system based on machine vision,which detects the defects of porous,less hole,hole offset,area deviation hole and edge burr hole in steel mesh.The main research contents are as follows:The research status of machine vision and automatic optical inspection technology at home and abroad was analyzed.The detection scheme based on machine vision was determined.The camera,lens and light source used were analyzed and the model most suitable for this system was selected.The software system includes image preprocessing algorithm,image mosaic algorithm and defect detection algorithm.For the noise existing in the steel mesh image,the median filtering is first used for noise reduction processing,and then the maximum inter-class variance method is used for threshold segmentation to extract the opening area and the connected domain mark is used to separate the different openings.Since the steel mesh is much larger than the lens field of view,it is necessary to splicing the collected steel mesh images,using the stitching method based on physical coordinates and the stitching method based on feature points.The splicing method based on physical coordinates mainly depends on the motion accuracy of the machine,and can meet the needs in most cases.Based on the feature point-based splicing method,the template matching method is added to the traditional feature point matching method,which solves the problem of mismatching in feature point matching.At the same time,it can overcome the influence of the machine position error on splicing and greatly improved the precision of the stitching.The defect detection algorithm uses a comparative reference method that is compared with the design file.The algorithm includes preliminary registration based on the minimum outer rectangle and secondary registration based on template matching to detect several common defects in the steel mesh.After that,the defect detection system was tested with a calibration plate and several steel meshes,and the existing problems were analyzed.Experiments have shown that the accuracy and efficiency of the system can meet today's industrial needs.