Design And Application Of Integrated Steel Mesh Storage And Inspection Syste

With the rapid development of information technology,the density of electronic products is increasing,and the open apertures of stencils in the production line of Surface Mounted Technology(SMT)are becoming more and more dense,with an increasing number of models and types.Therefore,the workload of stencil management is gradually expanding.If the residual solder paste on stencils is not detected and cleaned before placing them in the warehouse,it will gradually solidify in the storage process of the stencils.Even if these problems are discovered later,it is also difficult to completely remove the solder paste,which will eventually lead to the reduced service life of stencils.This paper studies and analyzes existing technologies at home and abroad and designs an integrated system for stencils that combines management and inspection functions.The main work is as follows:(1)Based on domestic GD32 microcontrollers,the hardware design of stencil management system is completed,including stencil management equipment,tension detection equipment and visual detection equipment.The hardware structure of each part is analyzed in detail,and the circuit and printed circuit board(PCB)are designed to construct the hardware equipment environment.Controller area network(CAN)is used to realize the interconnection and interworking of the internal components of the stencil management equipment.Bluetooth communication is used to realize the wireless transmission of tension data.Black-and-white industrial camera is used to build the image acquisition environment.(2)Defect detection methods based on machine vision are designed in order to realize the pre-inspection of stencils to screen out those with solder paste residues,including traditional image processing method and deep learning method.In the traditional image processing method,stencils are detected by reference and non-reference methods and quantified by comparison.In the deep learning method,the improved Ghost Net model is used for iterative training and verification based on the self-built stencil dataset,and finally the optimal model is obtained for classification and identification of stencils.(3)Some development tools are used to design the computer terminal software,the communication protocol between upper and lower computers,and the embedded device software,such as Py Qt and Keil.The hardware driver is implemented through the embedded device software,and human-computer interaction is achieved through the computer terminal software,which is integrated with the defect detection algorithm.The computer terminal software and embedded device software are interconnected based on the communication protocol between upper and lower computers.(4)A comprehensive test of the integrated system for stencils that combines management and inspection functions is carried out.The results show that the system can better realize the storage management of stencils,and the visual detection algorithm can meet the actual needs of defect detection for residual solder paste.The system functions normally and runs smoothly.