基于在线监测技术的铁路信号灯运行异常自动检测研究  

Research on automatic detection of abnormal operation of railway signal lights based on online monitoring technology

作　　者：段爱科 DUAN Aike(Lanzhou Raileay Design Institute Co Ltd.,Gansu Lanzhou 73000,China)

机构地区：[1]兰州铁道设计院有限公司,兰州730000

出　　处：《自动化与仪器仪表》2020年第12期46-49,共4页Automation & Instrumentation

基　　金：基于雷达切换的便携式双向作业防护装置(No.院科16-47)。

摘　　要：使用基于RGB颜色空间的阈值分割方法和Sobel算子受到铁路复杂环境影响,导致检测准确率较低,为解决铁路运输安全问题,对基于在线监测技术的铁路信号灯运行异常自动检测研究。将RGB颜色空间各分量转换为HSV空间模式,避免受到复杂环境干扰,统计目标全部像素点总和,利用点和线对偶特性,将HSV空间曲线转换成参数空间形式,使用Hough变换确定圆形度,以此进行铁路信号灯颜色特征提取。计算像素坐标横向和纵向梯度,得到梯度值,确定边缘法向的象限,完成图像边缘检测。检测亮的信号灯颜色,记录相关数据,由此设计自动检测流程。调试实验平台,避免实验过程受到外界环境干扰,由此进行实验验证分析。通过实验结果可知,该方法最高检测准确率为0.97,是提高铁路安全运输的有效方法。The threshold segmentation method based on RGB color space and the Sobel operator are affected by the complex environment of the railway,resulting in low detection accuracy.To solve the problem of railway transportation safety,automatic detection of abnormal operation of railway signal lights based on online monitoring technology is studied.The RGB color space components are corrverted to HSV space mode to avoid interference from complex environments.Counting the sum of all pixels of the target,the point and line dual characteristics are used to convert the HSV space curve into a parameter space form.Hough transform is used to determine the circularity In this way,railway signal light color feature extraction is performed.The horizontal and vertical gradients of the pixel coordinates are calculated toget the gradient value,determine the edge normal quadrant,and complete the image edge detection.The color of the bright signal light is deted and recorded the relevant data,thus the automatic detection process is designed.The experimental platform is debugged to avoid the interference of the external environment during the experimental process and then conduct the experimental verification analysis.According to the experimental results,the highest detection accuracy of this method is 0.97,which is an effective method to improve the safe transportation of railways.

关 键 词：在线监测技术 铁路信号灯 运行异常 自动检测 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]