The Design And Application Research Of Dimples Measurement And Classification System In Steel

In each kind of steel and iron material’s break failure analysis,we may understand that the breaking-down process the mechanism and the reason, subsequently verify enhance the materials mechanics performance the influencing factor by the fracture organization’s microscopic analytical study.Therefore,the qualitative and the quantitative analysis for fracture have the important fundamental research significance and the project application value.However,at present we can only depend upon the specialists to analyze dimple by the manual survey statistics pattern,and so the precision as well as the efficiency is low inevitably,and serious influence following material quality control work.Making full use of the achievements of primary researches and combining the present situation of our country digital image processing technology in the quota metallography,this paper,financed by the State Natural Science Foundation project and the Jiangsu University mold scientific innovation project,Has established the survey taxonomy system of dimple for the first time.This system mainly used the image processing way of filter,gray level amendment,division and so on to process the image of dimple,and utilized the morphology way of corrode methods and inflation to repair dimple.And the system use the way of improved based on the Freeman chain code multi-region area algorithm area of measurement to measure area. Main contents and innovations,1.First we analyze some kinds of algorithms of image processing,and obtain the suitable method on the base to analyze dimple characters.Namely we apply median filtering method to process image to eliminated effects of all kinds of noise in the image and modify the gray scale of picture to increase its contrast.2.In paper we introduced an adaptive method for image thresholding based on index of fuzziness to divide dimple from background,and eliminated the fuzzy characteristic which in the nest image existed,simultaneously retained the dimple edge feature.We used the morphology method and the scan line filling algorithm to solve the defects which exist in the dimple image.3.According to the dimple characteristic,this article used an improved multiregional area algorithm based on Freeman chain was presented to measure dimple.The results showed that the system has good effective for the processing of dimple images.With the above studies,the research is to provide essential theoretical evidence and technological basis for the further development of dimple quantitative classification analysis work.