现代科学仪器Issue(2)：24-28,5.
基于数字图像处理和特征提取的钢轨表面缺陷识别方法
Rail Defect Identification Using Digital Image Processing and Characteristics Extraction
摘要
Abstract
A rail surface defect identification method based on image characterize analysis is designed in this paper.The defect extraction of surface cracks and peeling pieces can be realized by using rapid digital image processing technology.During the identification process,top-hat and bottom-hat transform of sampled images is used for image enhancement;then Gaussian smoothing is adopted to eliminate small burrs;finally defect information is extracted through threshold processing and boundary inhibition.Rail defect evaluation and information representation methods are studied to present the defect information in the form of curves.In order to verify the validity of the method designed,an experimental system is developed,which uses industrial linear CCD to capture the images of a rail sample detected in motion and then identifies defects of the sample.Experiments showed that the information of defect level and location can be clearly expressed by the system’s output curves.关键词
钢轨探伤/高帽变换/低帽变换/图像增强/阈值处理/缺陷识别Key words
Rail defect detection/Top-hat/Bottom-hat/Information extraction and fusion/Threshold processing/Defect identification分类
机械制造引用本文复制引用
王平,刘泽,王嵬,张广伟,孙秀芳..基于数字图像处理和特征提取的钢轨表面缺陷识别方法[J].现代科学仪器,2012,(2):24-28,5.基金项目
国家自然基金资助项目 （）
北京市属高等学校人才强教计划资助项目 （）
