Depth Measurement for a Mobile Robot using Camera Rotation
Yu-Yuan Tsou, Huei‐Yung Lin · 2018
This paper presents an approach for depth measurement using a single camera with off-axis rotation. A camera is mounted on a rotational platform with the optical axis parallel to the rotation axis. The images captured from different viewpoints with visual parallax are obtained by camera circular motion with a baseline offset. Stereo matching is carried out on the image pairs rectified with camera calibration parameters. Depth computation is then performed using the disparity map and image coplanar geometry. Since multi-view images can be acquired with multiple camera rotation angles, the proposed method is able to reduce the correspondence mismatch in the multiple baseline stereo fashion. Experiments are carried out on the static and dynamic scene images captured using a mobile robot. The rudimentary results have demonstrated the feasibility of the proposed technique.