基于改进二维卷积滤波技术的轨道伤损检测方法
贾强韩峰
Track Damage Detection Method Based on Improved Two-dimensional Convolution Filtering Technology
JIA QiangHAN Feng
-
作者信息:兰州交通大学土木工程学院， 730070， 兰州
-
Affiliation:College of Civil Engineering, Lanzhou Jiaotong University, 730070, Lanzhou, China
-
关键词:
-
Key words:
-
DOI:10.16037/j.1007-869x.2022.04.039
-
中图分类号/CLCN:U4213.4+3
-
栏目/Col:应用技术
摘要:
基于改进卷积滤波技术和自适应阈值分割法，提出边缘检测轨道结构识别与伤损检测方法。使用改进二维卷积进行图像滤波,对原图像矩阵通过行列式变换处理，增强图像边界元素；根据边缘检测所得到的线形特征进行扣件弹条及钢轨识别，以线形急剧变化判别钢轨伤损及位置。通过既有线路轨道进行图像识别试验验证。试验验证结果表明，采用该方法可提高轨道部件识别速率，具有一定的自适应性，还可确定伤损相应位置，有利于轨道伤损的判别，可应用在日常轨道的养护及维修中。
Abstracts:
Based on improved convolution filtering technology and adaptive thresholding method, an image edge detection track structure recognition and damage detection method is proposed. Improved twodimensional convolution is used for image filtering, and the original image matrix is processed by determinant transformation to enhance the image boundary elements. According to the linear features obtained by edge detection, fastener springs and rails are identified, and drastic linear changes are used to identify rail damage and location. Image recognition test of existing line tracks is carried out for verification. Test results show that this method can raise the track component recognition rate and has certain adaptability, as well as to allocate the corresponding position of the damage, which is conducive to track damage identification and can be implemented in daily maintenance and repair of the track.
引文 / Ref:
贾强, 韩峰. 基于改进二维卷积滤波技术的轨道伤损检测方法. 城市轨道交通研究, 2022, 25(4): 181.JIA Qiang, HAN Feng. Track damage detection method based on improved two-dimensional convolution filtering technology. Urban Mass Transit, 2022, 25(4): 181.
- 上一篇: 高铁物流运输模式及其可行性
- 下一篇: 城市轨道交通运营安全的第三方评估
过刊浏览
资料下载
友情链接