Research On Edge Extraction And Evaluation For Shield Cutter Image

As an important tool in underground construction,the quality and working condition of shield cutters has a great impact on engineering efficiency and engineering quality.Because of the double effects of force and heat,the wear of shield cutters is unavoidable.If the wear condition of the shield cutters cannot be detected,it will not only cause delays in the construction period,but also cause damage to the mechanical equipment.In severe cases,it will reduce the service life of the equipment and even cause incalculable economic losses.A very important evaluation index for the wear status is the edge feature information of the shield cutters.Therefore,detecting the edge characteristics of shield cutters through related research methods and techniques to obtain their wear status is the key to ensure the normal operation of mechanical equipment,as well as the key to ensure engineering efficiency and quality.This thesis mainly focuses on the shield cutter images,and studies a more effective method of shield cutter edge extraction and evaluation based on image processing.For the uneven illumination and abrupt brightness in the working environment,this thesis studies the outer edge extraction method of the cutters to obtain more accurate image edge information,and evaluate the results.The main research contents of this thesis include:(1)Firstly,based on the image threshold segmentation algorithm,the target area and background area of the shield cutter image is segmented.Then,based on the edge detection operator,the cutter edge information is extracted to form the initial pixel edge pixel coordinate representation.Through analyzing the results of the contrast experiments,OTSU threshold segmentation and Canny edge detection operator is adopted.(2)A suitable curve fitting method to fit the edge contours is studied based on the cutter edge pixel coordinate information.And a continuous edge representation is built.Then,using the fitted edge contour as the heuristic path information,a rotation cutting method along the elliptical trajectory is designed.Starting from a point at an angle of 7.5 ° from the horizontal direction,the appropriate rotation angle and cutting window is set.Then the image is cut according to the elliptical trajectory.(3)We study an edge tracking algorithm for small areas.A new edge detection template for the Sobel operator is designed to extract the image gradient amplitude information.Based on the heuristic search algorithm,the image gradient value is used as the heuristic information to guide the extraction of the edge position information of the cutter in order to obtain more accurate external edge feature of the cutter image.(4)We study an evaluation method for the edge extraction quality of cutter images.The effectiveness of the edge extraction method is verified based on heuristic search from different perspectives.Firstly,the visual effect is verified intuitively.Secondly,the continuity effect of cutter image edge is verified by the quality evaluation based on connected components.Finally,the smoothness of cutter image edge is verified by the evaluation method based on distance variance.Compared with the traditional edge detection methods,the experimental results show that the edge detection algorithm based on heuristic search can effectively improve the continuity and smoothness of cutter image edge.