基于机器视觉的轨道异物侵限检测方法研究

Study on Intrusive Foreign Materials upon Railway Based on Bachine Vision

提出了一种快速的轨道异物检测及分割方法.首先是在图像中应用快速、准确的Hough变换检测出铁轨,再以铁轨图像参数确定工作窗口以检测并分割出异物.避免了对整幅图像内不必要的物体进行检测,有效提高了检测速度.通过实验证明该方法的实时性与有效性.

The paper proposes a method for quick detection and segmentation of railway foreign materials.Firstly quick and accurate Hough transformation is used to detect railway.Then work window is determined by the parameters of railway image to detect and segment foreign materials without detection of unwanted subjects.The detection speed is effectively raised.The experiments proves the real-time and validity of the proposed method.