基于改进Sobel算子的铁路轨枕自动定位方法  

Railway Sleeper Automatic Positioning Method Based on Improved Sobel Operator

作　　者：张天龙 许贵阳 ZHANG Tianlong;XU Guiyang(School of Mechanical-electronic and Vehicle Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China;Urban Rail Transit Vehicle Service Performance Guarantee Key Laboratory of Beijing,Beijing University of Civil Engineering and Architecture,Beijing 100044,China)

机构地区：[1]北京建筑大学机电与车辆工程学院,北京100044 [2]北京建筑大学城市轨道交通车辆服役性能保障北京市重点实验室,北京100044

出　　处：《铁道建筑》2019年第10期123-126,共4页Railway Engineering

基　　金：国家自然科学基金(51605023)

摘　　要：针对传统方法对铁路轨枕定位效率差、效果不理想等问题,提出一种基于改进Sobel算子的轨枕精确定位方法。首先将传统Sobel算子由原来水平和垂直2个方向模板增加至8个方向模板,提高边缘的检测精度;然后采用最大后验概率估计法对采用8方向Sobel算子检测出的梯度图像选取最佳阈值并进行二值化处理,增强算法的抗噪声能力;最后,将处理后的图像进行灰度投影,实现轨枕精确定位。试验结果表明,该方法能快速、准确地定位轨枕,为铁路轨枕状态检测提供可靠的、高精度的轨枕图像。Aiming at the problems of poor positioning efficiencies and unsatisfactory effects of traditional methods,amethod for accurate positioning of sleeper with improved Sobel operator was proposed. Firstly,the traditional Sobeloperator is increased from the original horizontal and vertical templates to 8-direction templates to improve the detectionaccuracy of the edges. Secondly,the maximum posterior probability estimation method is used to select best thresholdfor the gradient image detected by the 8-direction Sobel operator,and binarization is carried out to enhance the antinoise ability of the algorithm. Finally,gray projection is been done for processed image to achieve accurate positioningof the sleeper. The experimental results show that the method can position the sleeper rapidly and accurately,andprovide a reliable and accurate sleeper image for the subsequent sleeper state detection.

关 键 词：轨枕定位 现场试验 8方向Sobel算子 二值化 灰度投影 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] U213.3[自动化与计算机技术—控制科学与工程]