A Quality Inspection System Of Circular Blade Based On Machine Vision

In the process of modern industrial production,the requirements for production quality are higher and higher,especially in the production of lithium batteries.The quality of cutting tool is particularly important in the process of lithium battery production.Defects in the cutting tool can cause wavy edges,burrs,powder loss and other problems on lithium battery electrodes.In order to ensure the high quality production,this paper research the circular cutting tool quality inspection system.The system can effectively eliminate the traditional microscopic observation and detection methods of high damage to human eyes,poor accuracy,slow speed,poor repeatability and results are not easy to statistics.The machine vision processing technology is used to detect the quality of the round lithium battery electrode blade efficiently and accurately.According to the detective requirements of tool quality,this paper has made in depth research in hardware design,imaging design,algorithm design,accuracy verification,software design and other aspects:(1)According to the detective requirement of circular tool,this paper designs the hardware scheme and software scheme of online tool cutting edge testing system.And built a rotating platform,focusing and tracking system.According to the defect type of the cutting tool,I using a way that the camera and light source are on different sides,and I calibrated the image with the physical dimensions.(2)The automatic focusing system is designed to ensure fitting in different tool models.The contrast experiment of the commonly used image sharpness function is carried out,and the improved image sharpness function for the blade image is proposed.Finally,the improved peak search algorithm have been put forward to increase system`s focusing speed(3)In order to solve the problem of limited imaging field and difficult splicing,I proposed that the detection ROI area is variable.At the same time,in order to reduce the computational complexity,this paper first discusses the shortcomings of common curve similarity algorithm in tool contour matching,and then proposes an improved line similarity contour splicing and positioning algorithm,and verifies the reliability of this algorithm.(4)To solve the difficulties of high precision detection,this paper deeply studies the gauss sub-pixel and polynomial interpolation and Zernike algorithm.The edge profile was reconstructed by using the arc fitting method of constraint term.Then,the dust defect interference during contour reconstruction is optimized.Finally,the accuracy of the algorithm is verified by experiments.Finally,this paper has completed the construction of the experimental machine and software platform.plenty of experiments show that the algorithm proposed in this paper performs well in the detection of circular tool defects.And through the practical application in enterprises,it is show that this system not only has strong anti-jamming capacity and high detection accuracy,but also meets the demand of real-time detection,and can effectively reduce the burden of workers.