Research And Application Of Edge Detection Methods

With the development of electromechanical integration,parts of the large quantities of production requirements continue to increase.High-speed,high-precision infiltration to the processing of various production processes.However,in the processing process is often due to the outside world and the impact of the material itself,resulting in the production of the pin there are surface scratches,rust,size,and other defects beyond the scope of work,which will affect the normal use of the pin.For smaller pins,manual sorting of qualified parts has been unable to meet the industrial production of product quality and production speed requirements.Therefore,it is very important to study the technology of non-contact detection and screening of pins based on image processing.Through the study of many edge algorithms,the interpolation edge detection algorithm in the accurate positioning of the pin edge,in the actual image detection process,there is the disadvantage of inaccurate positioning.In this paper,the edge detection step is added to the pin selection process,and the original interpolation algorithm is optimized to solve the shortcomings of the theoretical algorithm.Based on the MATLAB programming platform,this thesis realizes the algorithm.The similarity function and Gaussian sub-pixel interpolation algorithm of SSIM are optimized in the process of pin surface quality and edge positioning.The algorithm is implemented by the corresponding program.The process is as follows:Firstly,the image is processed by Canny algorithm to get the image edge contour,and the outline is similar to the standard pin.Then,after the image Gaussian interpolation calculation,the exact size of the pin for screening.The combination of the two edge detection algorithms,in the process of edge detection with direction invariance,but also the efficiency of the pin to screen.Experiments show that the algorithm can be applied to pin edge extraction to meet the accuracy requirements...