Abstract: | This paper proposes the method detecting and tracking ships using navigational image sequence taken by the camera installed at ship's bridge. Since an image is influenced by roll and pitch, general image processing technique, such as Background Difference method cannot be applied. Moreover in a navigational image, the sea wave appears as noise. In this paper, we propose a novel technique to eliminate the influence of roll, pitch and sea waves in an image. And we tracked moving ships through navigational image sequence. The image is divided into small regions using brightness value, and these regions are labeled. Each of these regions is considered as a template. The template is assumed to be a ship. Then the template is matched on frames after selected time from an original image. During matching process, the coordinate points are recorded on an Excel table. The moving vector of a region is calculated on an Excel table. We distinguish ships from the character of a moving vector. The detected regions of ships are tagged. Each of the tagged regions is considered as a template and matched on frames after selected time. At this time, the template is replaced by new region to take measures against the change of illumination. The detection processes is started at a suitable interval to prepare for the appearance of another ships. Disappearances of ships are decided by the degree of coincidence of matching. This is the measure to against an occlusion or a frame out of ships. Twelve hours duration of video image have been taken in Tokyo Bay and Tokyo Port, and have been processed by proposal method. It is necessary to extract ship's regions precisely for exact tracking. For that, it is necessary to raise accuracy of extraction. We tried some method to enhance the accuracy of extraction and to prevent incorrect extraction. The result of an effective of measure has described in this paper. The video image that has been taken at Uraga Traffic Route captured several ships. Our method has been able to detect and track each ships even they have an occlusion. |
Published in: |
Proceedings of the 2006 National Technical Meeting of The Institute of Navigation January 18 - 20, 2006 Hyatt Regency Hotel Monterey, CA |
Pages: | 224 - 235 |
Cite this article: | Shimpo, M., Hirasawa, M., Yu, X., Arvelyna, Y., Oshima, M., "Method for Detecting and Tracking Ships Using Navigational Image Sequence as an Application of Integrated Navigation Systems," Proceedings of the 2006 National Technical Meeting of The Institute of Navigation, Monterey, CA, January 2006, pp. 224-235. |
Full Paper: |
ION Members/Non-Members: 1 Download Credit
Sign In |