A Visual Saliency Based Railway Intrusion Detection Method by UAV Remote Sensing Image

Railway intrusion is an important factor affecting the safety of train operations. The detection of railway intrusions based on the remote sensing image of low altitude UAV is a new and promising technology. This paper presents an intrusion detection method of UAV remote sensing image based on visual saliency. This method consists of four stages. First, the positions of the rails are fitted by differential excitation and Hough transform method and then the region of interest is extracted. Next, perspective transformation is applied to obtain the birds-eye view image. Then, motivated by the human visual attention mechanism, the salient areas are extracted based on the Hypercomplex Fourier Transform visual saliency detection and are considered as intrusions. Finally, the threshold method and morphological processing are used to extract the areas of intrusions. Experiments show that the proposed method can effectively detect and segment the areas of intrusion.