Research On Extracting Sea-Sky-Line And Target Tracking Algorithm Of Sea-Sky Background

Sea-sky-line detection is an important component involved in long-distance target tracking under the condition of complex sea-sky background. The stability of target tracking is very difficult in complicate sea-sky background because of the clutter caused by waves and clouds. So we can locate the sea-sky-line first, and then take the corresponding measures to reduce the complexity of the target tracking, so as to suppress the interference of wave, as well as improve the stability of target tracking and reduce the rate of false alarm.The characteristics of sea-sky background images is analyzed in this paper first, and many methods of extracting are studied, and then in the view of the difficulty to extract sea-sky-line in low-contrast images, an algorithm is presented which we called Hough transform based on phase grouping. As the algorithm combines the advantages of both the phase grouping and the Hough transform, which can not only extract the sea-sky-line under low-contract, but also has strong anti-interference, and overcome their own shortcomings of phase grouping and Hough transform.Because of the diversity of ship target images under complicated backgrounds, if the window of median filter is too small, the noise cannot be removed well; if the window is too large, the median filter cannot guarantee that the gray scale information of other objects in filtered image is reserved, and the following process cannot be executed smoothly. So two algorithms based on background complexity are proposed according to the feature that the sky and the sea usually show up at different gray values in sea-sky background images. The experiment results indicate that this method can extract sea-sky-line under simple and complicated sea-sky backgrounds, which further improves the adaptability of the algorithm.A large number of experiments show that when dealing with the multi-pixels images with the previous three methods, if the contrast of images is obvious they can work well; if not, they will fail. In one word single preprocessing method cannot accommodate all background images. Therefore, the complexity of background which can make a quantitative description of background complexity of background images, then images with different backgrounds can be treated differently based on this indicator, and real-time adjustments on preprocessing method can be made so as to realize the adaptive preprocessing of background images.The technology of target tracking under sea-sky background is also studied, and finally selects the centroid tracking algorithm. When the target is lost shortly, Kalman filter is used to predict the target trajectory, and solve the problem of interference to the stability of target tracking caused by waves.