摘要
针对现有钢轨伤损视觉检测未对伤损进行分类处理、缺乏针对钢轨接触疲劳裂纹的检测的问题,根据钢轨接触疲劳裂纹的方向性特征,提出了一种基于机器视觉的钢轨接触疲劳裂纹检测方法。首先利用纵向灰度值积分的方法从轨道图像中提取轨面区域;再将钢轨的图像划分为若干个单元区域,统计每个单元区域的梯度方向直方图;然后根据梯度方向特性判断单元区域是否存在接触疲劳裂纹。试验结果表明:该方法能有效地检测钢轨接触疲劳裂纹区域,并能排除光照不均等因素的干扰,检测的准确率达到97.33%。
Aiming at the problem that the existing visual inspection of rail damage usually does not deal with the damage types separately and there is a lack of inspection of rail contact fatigue cracks,a rail contact fatigue crack inspection method based on machine vision was proposed according to the directional characteristics of rail contact fatigue crack. Firstly,the track surface region was extracted from the track image by using the longitudinal gray value integration method. Then the rail image was divided into several unit regions,and the gradient direction histogram of each unit region was counted. Then,according to the gradient direction characteristics,whether there is contact fatigue crack in the unit region was judged. The test results show that this method can effectively inspect the rail contact fatigue crack area,eliminate the interference of uneven illumination,and the inspection accuracy is 97.33%.
作者
张博
刘秀波
ZHANG Bo;LIU Xiubo(Infrastructure Inspection Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)
出处
《铁道建筑》
北大核心
2021年第11期116-119,共4页
Railway Engineering
基金
中国铁道科学研究院集团有限公司基金(2018YJ146)。
关键词
机器视觉
钢轨接触疲劳裂纹
伤损检测
直方图
梯度方向筛选
machine vision
rail contact fatigue crack
damage inspection
histogram
gradient direction screening
