| 基于改进马尔可夫随机场的钢轨缺陷分割 |
| |
| 引用本文： | 张辉,李平,Q.M.Jonathan WU,贺振东.基于改进马尔可夫随机场的钢轨缺陷分割[J].计算机工程与设计,2020,41(4):1052-1061. |
| |
| 作者姓名： | 张辉 李平 Q.M.Jonathan WU 贺振东 |
| |
| 作者单位： | 长沙理工大学电气与信息工程学院,湖南长沙 410114;温莎大学电气与计算机工程系,加拿大温莎 N9B3P4;长沙理工大学电气与信息工程学院,湖南长沙 410114;温莎大学电气与计算机工程系,加拿大温莎 N9B3P4;郑州轻工业学院电气信息工程学院,河南郑州450002 |
| |
| 基金项目： | 国家自然科学基金;长沙市科技计划;湖南省教育厅科学研究项目;国家科技支撑计划 |
| |
| 摘 要： | 为提高钢轨缺陷分割对噪声的鲁棒性,提出一种基于改进马尔可夫随机场(MRF)的钢轨缺陷分割方法。利用背景差分法对灰度进行预处理,消除灰度分布不均的干扰。对模糊if-then规则的前提部分采用马尔可夫随机场来利用图像中的空间约束,结果部分指定像素距离图算法,通过使用马尔可夫随机场(MRF)在相邻像素图像之间并入局部空间信息,推导出新的自适应模糊集和MRF相结合的钢轨表面缺陷自动分割方法。建立标准的FCM、GMM和该方法的钢轨缺陷分割对比实验,验证了算法的有效性和优越性。
|
| 关 键 词： | 钢轨缺陷 背景差分 马尔可夫随机场 空间信息 缺陷分割 |
Rail defect segmentation based on improved Markov random field |
| |
| ZHANG Hui,LI Ping,WU QMJonathan,HE Zhen-dong.Rail defect segmentation based on improved Markov random field[J].Computer Engineering and Design,2020,41(4):1052-1061. |
| |
| Authors: | ZHANG Hui LI Ping WU QMJonathan HE Zhen-dong |
| |
| Affiliation: | (School of Electrical and Information Engineering,Changsha University of Science and Technology,Changsha 410114,China;Department of Electrical and Computer Engineering,University of Windsor,Windsor N9B3P4,Canada;School of Electrical and Information Engineering,Zhengzhou University of Light Industry,Zhengzhou 450002,China) |
| |
| Abstract: | To improve the robustness of rail defect segmentation to noise,a rail defect segmentation method based on improved Markov random field(MRF)was proposed.The background difference method was used to pre-process the gray level,eliminating the interference of uneven distribution of gray level.For the premise part of the fuzzy IF-THEN rule,Markov random field was used to utilize the spatial constraints in the image,and the result part specified the distance map algorithm of the pixels.By incorporating local spatial information between adjacent pixel images using Markov random field(MRF),a new automatic rail surface defect segmentation method combining adaptive fuzzy set and MRF was derived.A standard FCM,GMM and a comparative experiment of rail defect segmentation based on this method were established to verify the effectiveness and superiority of the algorithm. |
| |
| Keywords: | rail surface defect background difference Markov random field spatial information defect segmentation |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
