단일 카메라를 이용한 수중 로봇의 자세 추정 법: 자율 도킹 방법으로써 응용 가능성 고찰

Abstract

In this paper, we propose a complete vision based solution of a pose estimation method for the autonomous underwater docking problem. The proposed method searches for the corner points of the checker board pattern. These corners are accurately labeled with respect to the gray box region of the pattern board in order to find a homography relation between the pattern coordinate frame and the camera coordinate frame. Each homography is used for estimating the extrinsic parameters of the camera so as to determine the optimal path or sequence of the docking task. The proposed method is validated through real underwater image data set The obtained results indicate that the proposed method is quite feasible for the autonomous docking task for underwater robots.