Research Of Gear Image Boundary Extraction Algorithm Based On Machine Vision Measurement

Gear is a basic part of motion and power transmission.While the gear manufacturing accuracy directly affects the quality of the work.Therefore,gear measurement has become a very important link in the gear studying and production.Traditional contact gear measurement method has low precision,heavy workload and other defects,so measurement method based on machine vision appeared.While the image edge extraction is the premise for the subsequent image processing and the gear parameters taking.Gear measurement based on machine vision is the research background,machine vision technology and digital image processing technology is the theoretical basis,the paper puts forward a new edge extraction algorithm to the back-light straight tooth gear image.The main work is as follows:First,this paper discusses the importance of the gear precision measurement and the feasibility and necessity of machine vision technology in gear detection.While gear image edge detection is the necessary premise for gear measurement.Thus,through the study of the edge,we can see that the ideal edge mainly consists of two kinds: step type and ridge type.Then,the edge types of the gear images are analyzed.Second,several classical pixel level edge detection algorithms are studied,and these operators are used as contrast experiments.The advantages and disadvantages of these algorithms are summarized by combining the experimental results with the theoretical basis.Based on the above analysis,put forward pixel edge extraction algorithm based on eight neighborhood search.Based on the location relation of pixel eight neighborhood,the image is smoothed by Gauss filter,and the edge is extracted according to the relation of the gray value between the target pixel and the eight neighborhood pixel.Third,the existing sub-pixel edge detection algorithms are studied,including fitting method,interpolation method and moment method.Through the research on the existing subpixel edge extraction algorithm,proposed a sub-pixel edge extraction algorithm based on the bi-linear interpolation and Gauss curve fitting,the experimental results show that the algorithm not only guarantees the edge precision,also can reduce the calculation time.