Citation: | Dai Yongshou, Liu Bowen, Li Ligang, et al. Sea-sky-line detection based on local Otsu segmentation and Hough transform[J]. Opto-Electronic Engineering, 2018, 45(7): 180039. doi: 10.12086/oee.2018.180039 |
[1] | 曾文静, 万磊, 张铁栋, 等.复杂海空背景下弱小目标的快速自动检测[J].光学精密工程, 2012, 20(2): 403–412. Zeng W J, Wan L, Zhang T D, et al. Fast detection of weak targets in complex sea-sky background[J]. Optics and Precision Engineering, 2012, 20(2): 403–412. |
[2] | Fefilatyev S, Goldgof D, Shreve M, et al. Detection and tracking of ships in open sea with rapidly moving buoy-mounted camera system[J]. Ocean Engineering, 2012, 54: 1–12. doi: 10.1016/j.oceaneng.2012.06.028 |
[3] | 王丁禾, 牛照东, 张聘义, 等.利用模糊综合评判技术提取红外图像海天线[J].光电工程, 2012, 39(11): 67–74. doi: 10.3969/j.issn.1003-501X.2012.11.011 Wang D H, Niu Z D, Zhang P Y, et al. Sea-sky-line extraction in infrared images using fuzzy synthetical judgment[J]. Opto-Electronic Engineering, 2012, 39(11): 67–74. doi: 10.3969/j.issn.1003-501X.2012.11.011 |
[4] | 王博, 苏玉民, 万磊, 等.基于梯度显著性的水面无人艇的海天线检测方法[J].光学学报, 2016, 36(5): 58–67. Wang B, Su Y M, Wan L, et al. Sea sky line detection method of unmanned surface vehicle based on gradient saliency[J]. Acta Optica Sinica, 2016, 36(5): 58–67. |
[5] | 徐良玉, 马录坤, 谢燮, 等.基于结构森林边缘检测和Hough变换的海天线检测[J].上海大学学报(自然科学版), 2017, 23(1): 47–55. Xu L Y, Ma L K, Xie X, et al. Sea-sky line detection based on structured forests edge detection and Hough transform[J]. Journal of Shanghai University (Natural Science), 2017, 23(1): 47–55. |
[6] | Mou X Z, Wang H. Image-based maritime obstacle detection using global sparsity potentials[J]. Journal of Information and Communication Convergence Engineering, 2016, 14(2): 129–135. doi: 10.6109/jicce.2016.14.2.129 |
[7] | Prasad D K, Rajan D, Prasath C K, et al. MSCM-LiFe: multi-scale cross modal linear feature for horizon detection in maritime images[C]//Proceedings of 2016 IEEE Region 10 Conference, Singapore, 2016: 1366–1370. |
[8] | 刘松涛, 周晓东, 王成刚.复杂海空背景下鲁棒的海天线检测算法研究[J].光电工程, 2006, 33(8): 5–10. Liu S T, Zhou X D, Wang C G. Robust sea-sky-line detection algorithm under complicated sea-sky background[J]. Opto-Electronic Engineering, 2006, 33(8): 5–10. |
[9] | Kristan M, Kenk V S, Kovačič S, et al. Fast image-based obstacle detection from unmanned surface vehicles[J]. IEEE Transactions on Cybernetics, 2016, 46(3): 641–654. doi: 10.1109/TCYB.2015.2412251 |
[10] | 韩嘉隆, 毛征, 王宁, 等.基于二维OTSU的海天分界线提取算法[J].国外电子测量技术, 2016, 35(8): 67–70. Han J L, Mao Z, Wang N, et al. Algorithm for sea-sky-line extraction based on two-dimension OTSU[J]. Foreign Electronic Measurement Technology, 2016, 35(8): 67–70. |
[11] | 谢红, 刘玲, 刘艳艳.复杂海天线区域检测算法研究[J].应用科技, 2006, 33(6): 96–98. Xie H, Liu L, Liu Y Y. Research on complicated sea-sky-line area detection algorithm[J]. Applied Science and Technology, 2006, 33(6): 96–98. |
[12] | 刘士建, 蒋敏, 庄良.一种快速有效的红外图像中海天线提取算法[J].红外技术, 2011, 33(4): 230–232, 240. Liu S J, Jiang M, Zhuang L. A fast and effective algorithm for sea-sky-line extraction in infrared images[J]. Infrared Technology, 2011, 33(4): 230–232, 240. |
[13] | 邹瑞滨, 史彩成, 毛二可.基于剪切波变换的复杂海面红外目标检测算法[J].仪器仪表学报, 2011, 32(5): 1103–1108. Zou R B, Shi C C, Mao E K. Shearlet-based infrared target detection algorithm on complex sea[J]. Chinese Journal of Scientific Instrument, 2011, 32(5): 1103–1108. |
[14] | Gonzalez R C, Woods R E. Digital Image Processing[M]. Ruan Q Q, Ruan Y Z, trans. 3rd ed. Beijing: Publishing House of Electronics Industry, 2017. |
[15] | Prasad D K, Rajan D, Rachmawati L, et al. Video processing from electro-optical sensors for object detection and tracking in a maritime environment: a survey[J]. IEEE Transactions on Intelligent Transportation Systems, 2017, 18(8): 1993–2016. doi: 10.1109/TITS.2016.2634580 |
[16] | Bloisi D D, Iocchi L, Pennisi A, et al. ARGOS-Venice boat classification[C]//Proceedings of the 12th IEEE International Conference on Advanced Video and Signal Based Surveillance, Karlsruhe, Germany, 2015: 1–6. |
Overview: Unmanned surface vehicle (USV) has a great potential to play an important role in the near future, such as sea environmental monitoring and maritime rescue. USV obtains information about surrounding sea surface environment by processing the visible light maritime image from the camera mounted on the USV. Sea-sky-line detection is useful in the visible light maritime image processing. It can provide important reference for the target detection and image calibration. Existing sea-sky-line detection methods are mainly used in infrared maritime images with simple scenes and less interference. In contrast, there are few studies on sea-sky-line detection in complex visible light maritime images. There are two main methods for the detection of sea-sky-line, namely the method based on line extraction from edge pixels and the method based on image segmentation. However, the former method is susceptible to the gradient change of sea waves and sea-sky-line, while the latter is limited by the accuracy of image segmentation. Due to the interference such as sea waves, ships and light, it is difficult to accurately detect the sea-sky-line of the visible light maritime image. To improve the detection accuracy and robustness, a sea-sky-line detection method based on local Otsu segmentation and Hough transform is proposed. Firstly, high-frequency noise such as light spot in the gray image is rapidly suppressed by longitudinal median filter. Then, according to the image features, local Otsu segmentation is performed to obtain binary images where edge pixels are extracted. Finally, Hough transform is used to fit edge pixels to complete the sea-sky-line detection. In the proposed method, image block processing compensates for the inhomogeneity of illumination and limits the interference scope of ships to some image blocks, which makes the local Otsu segmentation more accurate than the global Otsu segmentation. In addition, compared with the edge detection of the sea-sky-line based on the gradient, the edge detection of the sea-sky-line based on image segmentation can better adapt to the change of the image gradient and suppress the interference of the wave edge. Hough transform can ensure the accurate fitting of the sea-sky-line from the edge pixel if the number of edge pixels extracted of the sea-sky-line is more than half of the image width. Experimental results show that the interference of sea waves, ships and light can be effectively suppressed by the proposed method, which is relatively accurate, robust and real-time. The sea-sky-line detection accuracy of the proposed method is 93.0%, which is significantly higher than that of three representative sea-sky-line detection methods.
Flowchart of sea-sky-line detection based on local Otsu segmentation and Hough transform
Analysis and filter of high frequency noise such as light spot. (a) Original images; (b) The profile plot of the gray image single column; (c) The median filter result of Fig. 2(b)
Segmentation results comparison of typical maritime images. (a)~(d) Original images; (e)~(h) Results of global Otsu segmentation; (i)~(l) Results of local Otsu segmentation
Edge detection results comparison of typical maritime images. (a)~(d) Edge detection results of Canny with low gradient threshold; (e)~(h) Edge detection results of Canny with high gradient threshold; (i)~(l) Edge detection results based on local Otsu segmentation
Sea-sky-line detection results of the proposed method on typical maritime images. (a)~(d) Images with many waves; (e)~(h) Blurred images; (i)~(l) Images with sea reflection; (m)~(p) Images with close target
Sea-sky-line detection results comparison of three methods. (a)~(d) Results of Prasad method; (e)~(h) Results of Fefilatyev method; (i)~(l) Results of Kristan method
Failure cases. (a) The image that sea-sky line is shaded; (b) The local Otsu segmentation results of image 7(a); (c) The image that sea-sky region is blurred; (d) The local Otsu segmentation results of image 7(c)