Research On Detection Methods Of Abrasion On The Surface Of Cylinder Liner Based On Image Processing Technology

Cylinder liner is the main part of maritime shipping diesel engine, its operating well or not influences on the performance of the diesel engine. The paper systematically analyses and induces the styles of abnormal abrasion on the surface of cylinder liner, producing mechanism and the specific reasons of causing abnormal abrasion; on above basis, sets forth the features of abnormal abrasion in the image. It pays the way for the future research. The paper introduces image edge extraction technology in order to make a distinction between abrasive objectives and settings on the surface of abrasive image.The edge is the important feature that human vision recognizes image target. The main information of image that all consists in the edge is a sharper place of the gray scale change in the image, and shows the discontinuity of local characters in the image. How to extract the edge character of target that is needed is a difficult and hot spot in the image processing. Now there are many methods of edge extraction to have been researched, but none of them is an effective method that is suitable to all kinds of image, especially to the noise image. Firstly the paper analyzes features of digital images, then applies the technology of histogram equalization and median filtering to process enhancement of contrast ratio and imaging denoising of collected abrasive images on the surface of cylinder liner. On preprocessing abrasive images basis, from the classical methods of edge extraction, the paper, based on the image preprocessing of abrasion, studies two perfected methods of edge detection Marr-Hildreth and Canny, finally introduces the methods of multiscale edge detection based on wavelet transform and applies it to edge extraction of the abrasive image. According to the experiment, the method is an effective one for edge extraction of the abrasive image.The paper studies the application of template matching to different abrasive image recognition. Based on the edge feature of extracted abrasive image, the paper puts forward HD on the basis of the edge feature and makes the recognition of imagematching based on different abrasive styles in order to realize the rapid algorithm and to save operational time. According to matching results, the matching recognition ratio of edge character images based on different edge detection methods is different. On the whole, the method can effectively recognize different abrasive types of images on the surface of cylinder liner.The paper applies the technology of image processing and recognition to the detection of abnormal abrasion on the surface of cylinder liner. In view of several kinds of classical abnormal abrasive image, it uses the technology of digital image processing to extract, analyze and study abnormal image features by analyzing the mechanism of abrasion on the surface of cylinder liner, and realizes the recognition of template matching on the abrasive feature. It refers to maintenance of managerial personnel on shipping, and opens up a new method to detect the abrasion of cylinder liner.