摘要
铁路道岔是高速铁路系统最重要的线路连接设备,随着列车运行速度的提高,列车运行对道岔的冲击越来越大,因此对高速铁路道岔缺口的监测、维护和调整要求也逐渐提高。近年来,基于图像的方法在道岔缺口监测领域得到应用和推广。本文针对高速铁路道岔缺口的定位和监测,构建基于CMOS面阵的道岔缺口图像监测架构。利用信息论的方法,提出基于香农熵的自适应阈值图像边界分割算法,实现算法的低计算复杂度,适应不同对比度和噪声环境下的道岔缺口检测。利用京广铁路长沙站的图像数据,对算法性能进行分析对比。实验结果表明本文所提出的边界检测算法和道岔缺口监测架构具有一定的有效性。
Railway switch is the most important route connecting device in high-speed rail systems.With the increase of train speed,the impact of train operation on switches is increasing.Therefore,higher requirements for detecting,maintaining and tuning switch gaps have been put forward.In recent years,image-based methods have been applied to the field of switch gaps detection.In this research,a framework of CMOS plane based detection of switch gaps was constructed focusing on the location and monitoring of switch gaps for the high-speed railways.By means of the theory of information,Shannon entropy based adaptive threshold selection algorithm and edge detection algorithm were proposed to realize low computational complexity of the algorithm,to adapt to switch gap detection under various image contrasts and noise levels.The performance of the proposed framework and the edge detection algorithm was analyzed by image data from Changsha Station in Beijing-Guangzhou high-speed railway.The experimental results showed the efficiency of the proposed switch gap detection framework and the Shannon entropy based edge detection algorithm.
出处
《铁道学报》
EI
CAS
CSCD
北大核心
2016年第12期70-75,共6页
Journal of the China Railway Society
基金
中国铁路总公司科技研究开发计划(2014X008-A)