基于2D/3D成像的地铁列车螺栓松动检测研究与应用
Research and Application of Metro Train Bolt Loose Detection Based on 2D/3D Imaging
- 机车电传动 2021年0卷第6期 页码：147-154
- 作者机构：
1.广东广佛轨道交通有限公司，广东 佛山 528251
2.广州轨道交通建设监理有限公司，广东 广州 510010
3.广州运达智能科技有限公司，广东 广州 510445
- 作者简介：
王勇（1981—），男，高级工程师，主要研究方向为城轨车辆智能运维与运营管理；E-mail：rachel0105@126.com
- 基金信息：
DOI：10.13890/j.issn.1000-128x.2021.06.021
中图分类号：
扫 描 看 全 文
王勇, 邓卓森, 王志云, 等. 基于2D/3D成像的地铁列车螺栓松动检测研究与应用[J]. 机车电传动, 2021,0(6):147-154.
Yong WANG, Zhuosen DENG, Zhiyun WANG, et al. Research and Application of Metro Train Bolt Loose Detection Based on 2D/3D Imaging[J]. Electric Drive for Locomotives, 2021,0(6):147-154.
王勇, 邓卓森, 王志云, 等. 基于2D/3D成像的地铁列车螺栓松动检测研究与应用[J]. 机车电传动, 2021,0(6):147-154. DOI： 10.13890/j.issn.1000-128x.2021.06.021.
Yong WANG, Zhuosen DENG, Zhiyun WANG, et al. Research and Application of Metro Train Bolt Loose Detection Based on 2D/3D Imaging[J]. Electric Drive for Locomotives, 2021,0(6):147-154. DOI： 10.13890/j.issn.1000-128x.2021.06.021.
摘要
在地铁列车智能运维领域，列车外观异常检测是必不可少的重要环节，从传统的人工巡检到采用线阵相机2D成像图像识别方法，均存在一定的缺陷，不能满足快速准确的智能巡检要求。基于2D成像的图像识别方法可以解决人工巡检速度慢和因疲劳导致漏检等问题，但在与成像平面垂直的第三维度上出现的异常问题还没有得到较好的解决。文章提出一种基于2D/3D成像的螺栓松动检测方法，避免了在2D成像过程中由于光照、水渍、灰尘等原因造成图像像素值变化过大对检测结果带来的影响，并能很好地应用于地铁列车异常检测。
Abstract
In the field of intelligent operation and maintenance of subway trains, the detection of abnormal appearance of trains is an indispensable important link. From traditional manual inspections to 2D imaging image recognition methods of line array cameras, there are certain defects, which are not enough to achieve fast and accurate intelligent inspections. The image recognition method based on 2D imaging can solve the problems of slow manual inspection and missed inspection caused by fatigue. However, the abnormal problems that occur in the third dimension perpendicular to the imaging plane have not been well resolved. A bolt looseness detection method based on 2D/3D imaging was proposed, which avoided the impact of excessive changes in image pixel values caused by light, water stains, gray layers, etc. on the detection results in 2D imaging, and can be applied to abnormal detection of subway trains very well.
关键词
地铁3D检测平面拟合区域生长松动计算
Keywords
subway3D detectionplane fittingarea growthloose calculation
references
杨佳烁. 基于3D成像技术的铁路扣件检测研究与应用[D]. 北京: 北京交通大学, 2019.
YANG Jiashuo. The research and application of railway fasteners detection based on 3D imaging technology[D]. Beijing: Beijing Jiaotong University, 2019.
李静. 机车车底关键螺栓故障检测技术研究[D]. 成都: 西南交通大学, 2015.
LI Jing. Key locomotive bolts fault detection technique[D]. Chengdu: Southwest Jiaotong University, 2015.
宋鑫鑫. 基于图像处理技术的列车部件异常自动检测方法研究及应用[D]. 西安: 长安大学, 2019.
SONG Xinxin. Research and application of automatic detection method for train parts abnormal based on image processing technology[D]. Xi'an: Chang'an University, 2019.
张瑞娟, 蒲宝明, 潘世铭. 模板匹配算法在动车零部件故障检测系统中的应用[J]. 计算机系统应用, 2011, 20(8): 134-137.
ZHANG Ruijuan, PU Baoming, PAN Shiming. Template matching algorithm in the study of CRH train parts fault detection systems[J]. Computer Systems & Applications, 2011, 20(8): 134-137.
王宝丽. 基于深度学习的中低速磁浮F轨螺栓松动识别研究[D]. 成都: 西南交通大学, 2019.
WANG Baoli. Research on bolt loosening recognition of medium and low speed maglev F-rail based on deep learning[D]. Chengdu: Southwest Jiaotong University, 2019.
张宏伟, 赖百炼. 三维激光扫描技术特点及其应用前景[J]. 测绘通报, 2012(增刊1): 320-322.
ZHANG Hongwei, LAI Bailian. Characteristics and application prospect of three-dimensional laser scanning technology[J]. Bulletin of Surveying and Mapping, 2012(Suppl 1): 320-322.
SHELHAMER E, LONG Jan, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 39(4): 640-651.
REN Shaoqing, HE Kaiming, GIRSHICK Ross, et al. Faster R-CNN: Towards real-time object detection with region proposal networks[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 39(6): 1137-1149.
王峰, 丘广新, 程效军. 改进的鲁棒迭代最小二乘平面拟合算法[J]. 同济大学学报(自然科学版), 2011, 39(9): 1350-1354.
WANG Feng, QIU Guangxin, CHENG Xiaojun. An improved robust method for iterating least-squares plane fitting[J]. Journal of Tongji University(Natural Science), 2011, 39(9): 1350-1354.
FISCHLER M A, BOLLES R C. Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography[J]. Communications of the ACM, 1981, 24(6): 381-395.
Rahul R, CHUM O, POLLEFEYS M, et al. USAC: a universal framework for random sample consensus.[J]. IEEE transactions on pattern analysis and machine intelligence, 2013, 35(8): 2022-2038.
李孟迪, 蒋胜平, 王红平. 基于随机抽样一致性算法的稳健点云平面拟合方法[J]. 测绘科学, 2015, 40(1): 102-106.
LI Mengdi, JIANG Shengping, WANG Hongping. A RANSAC-based stable plane fitting method of point clouds[J]. Science of Surveying and Mapping, 2015, 40(1): 102-106.
何俊, 葛红, 王玉峰. 图像分割算法研究综述[J]. 计算机工程与科学, 2009, 31(12): 58-61.
HE Jun, GE Hong, WANG Yufeng. Survey on the methods of image segmentation research[J]. Computer Engineering and Science, 2009, 31(12): 58-61.
LIN Z, JIN J, TALBOT H. Unseeded region growing for 3D image segmentation[C]//VIP'00: Selected papers from the Pan-Sydney workshop on Visualisation. Sydney: [s.n], 2000(2): 31-37.
武新丹. 基于区域生长技术的三维模型分割研究与应用[D]. 太原: 中北大学, 2020.
WU Xindan. Research and application of 3D model segmentation based on region growth technique[D]. Taiyuan: North University of China, 2020.
0
浏览量
27
下载量
0
CSCD
3
CNKI被引量
- 网络PDF下载
- Meta-XML下载
- BibTeX下载
- Endnote下载
- RefMan 下载
- NoteExpress下载
- NoteFirst下载
关联资源
相关文章
相关作者
相关机构
- 技术支持由北京北大方正电子有限公司提供 湘ICP备20010475号-1
京公网安备11010802024621 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies