Abstract—Sea-sky-line/coastline extraction is a key problem in infrared maritime target search and track systems. However, the current studies of sea-sky-line/coastline detection in infrared imagery usually aim for detection accuracy or time saving unilaterally, and it is seldom suitable for practical applications in complex scenarios. To cope with this problem, a novel sea-sky-line/coastline extraction method based on the structure tensor and Hough transform is proposed in this paper. Firstly, the Gaussian filter based pre-processing is applied on the infrared image to eliminate the clutters. Then, the Gaussian structure tensor is proposed to enhance the line property of sea-sky-line/coastline and a segmentation threshold is selected to binarize the image. Finally, on the basis of line property of sea-sky boundary in binary image, the Hough transform is applied to effectively locate the sea-sky-line/coastline. Extensive experiments show that the proposed sea-sky-line/coastline extraction method not only has an outstanding detection accuracy under different scenarios but also has the potential of real-time hardware processing. 

Index Terms—sea-sky-line/coastline extraction, infrared imagery, structure tensor, Hough transform

Cite: Yongsong Li, Zhengzhou Li, Weiqi Xiong, and Jiangpeng Yu, "Sea-Sky-Line/Coastline Extraction in Infrared Imagery Based on Structure Tensor and Hough Transform," International Journal of Signal Processing Systems, Vol. 7, No. 4, pp. 123-128, December 2019. doi: 10.18178/ijsps.7.4.123-128