基于改进Canny算法的铁轨边缘检测  

Rail Edge Detection Based on Improved Canny Algorithm

作　　者：王燕芩 李沛奇[2] 

机构地区：[1]卡斯柯信号有限公司 [2]兰州铁路局

出　　处：《铁道通信信号》2015年第2期75-78,共4页Railway Signalling & Communication

摘　　要：针对异物入侵的边缘检测方法进行研究。根据中值滤波算法和开关滤波原理,结合现场复杂多变的监测环境,采用改进的开关型中值滤波器,以分段统计的方法对图像进行去噪处理。为了更好的恢复被破坏的数据,结合铁轨具有连续边缘走向的特点,利用惯性原理进行边缘追踪,提出改进的Canny算法,用于判断铁轨的异物入侵,并对问题进行了试验和分析。试验结果表明,该算法的噪声抑制能力远优于传统算法,能适合现场的监测,具有更好的连续性和细节边缘表现力,提高了整个监测系统的自动化程度和可靠性。An edge detection approach of railway foreign invasion is studied in the paper.Considering the complex and variable monitoring environment on site,an improved switched median filter is used to remove noise using a staged statistics method according to median filter algorithm and the principle of switched capacitor filter.In order to recover the damaged data,a method based on the principle of inertia is used to trace the edge according to the continuity of railway edge.Through analysis of experiment and test results,it demonstrates that the algorithm is fit for field monitoring and also can improve the automation and reliability of the system.

关 键 词：铁路运输 异物侵限 CANNY算子 开中值滤波器 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]