铁轨检测中基准桩的识别及其中心确定

The Recognition of the Reference Stake and Its Center Calculation in Track Inspection System

本文运用机器视觉的方法进行铁轨建设时基准桩的检测,克服了传统测量方法费时费力的缺点,提高了检测的速度和效率。在机器视觉检测系统中,利用工业CCD相机采集图像数据,用序贯相似性检测算法(SSDA)模板匹配法识别基准桩,通过合理选择阈值,准确地将基准桩分离出来。再通过Canny算子求得基准桩的边缘图像,利用边缘信息找出基准桩的中心,通过相机坐标系和成像坐标系间的换算,即可得到基准桩在测量坐标系中的位置。实验表明,这种方法快速、省力、精度较高,有良好的应用价值。

This paper inspects the reference stake in railway by employing the machine vision method,which improves inspecting velocity and efficiency corresponding to the traditional method.In the machine vision inspection system,the paper collects images through industrial CCD camera,recognizes the reference stake using SSDA template matching and finds the reference stake correctly by proper threshold.The paper utilizes the Canny operator to detect the edge of object and gives an algorithm of figuring out the ge-ome...