Method of Eliminating and Segmenting Ferrographic Images Nonuniform Background
For most of optical microscopes, the fact that a point light source is adopted results in the nonuniform distribution of illuminance. As a result, for the ferrography image gained using optical microscopy, the image pixeis are brighter in the centre than in the edge. Undoubtedly, this will influence significantly the precision of binary image segmentation which is a key step to gain the characterization of wear debris. Therefore, it is very imperative to eliminate the nonuniform distribution of the background brightness of the ferrography image before performing an effective and accurate binary image segmentation. This project is to investigate the elimination of the nonuniform distribution of the background brightness of the ferrographic images in order to gain the accurate binary image of the ferrographic image. Different methods based on the first-order gradient operators, the second-order gradient operators and compensation of the nonuniform distribution of the illumination are developed and their comparison are also presented in the paper. It is believed that the knowledge presented in this paper will be helpful to get the accurate binary image of the ferrographic image which is the foundation of analyzing morphology of wear debris.
Ferrographi Image Noise Image Segmentation
Lu Zhiyong Peng Qi Yan Xinping
Intelligent Transport System Research Center, Wuhan University of Technology Reliability Engineering Institute, Wuhan University of Technology Wuhan,hubei,430063, P.R.China
国际会议
长沙
英文
1765-1768
2010-03-13（万方平台首次上网日期，不代表论文的发表时间）