Syoji Kobashi, Hokuto Mita, Kazuya Nakagawa, Koji Nishiyama, Hitoshi Maeno, Kei Kuramoto, Yutaka Hata
Proceedings of the 2013 IEEE Workshop on Robotic Intelligence in Informationally Structured Space, RiiSS 2013 - 2013 IEEE Symposium Series on Computational Intelligence, SSCI 2013, 102-107, 2013 Peer-reviewed
Understanding a space where autonomous robots work is an open problem. In case of maritime space, it is also very important. For this purpose, marine radar has been used to acquire images around the maritime vehicles. However, radar images are easily distorted by signal attenuation with distance, blurring due to antenna directivity, reflection by obstructs, etc. There are some conventional methods to improve the image quality. However, there is a limitation due to the acquisition mechanism of radar systems. To overcome the limitation, this study shows a novel approach, which uses multiple radar images to understand the maritime space. The method estimates radar cross section (RCS) from multiple radar images by iterative image reconstruction algorithm. Performance of the proposed method is validated using actual radar images taken by marine radar system equipped on a ship. The experimental results showed that the proposed method presents a maritime space map with high image quality and without distance attenuation, sidelobe diffusion. © 2013 IEEE.