International Journal of Computer Vision
ICCV '98 Proceedings of the Sixth International Conference on Computer Vision
Efficient Dense Stereo with Occlusions for New View-Synthesis by Four-State Dynamic Programming
International Journal of Computer Vision
Stereo for slanted surfaces: first order disparities and normal consistency
EMMCVPR'05 Proceedings of the 5th international conference on Energy Minimization Methods in Computer Vision and Pattern Recognition
Hi-index | 0.00 |
In recent years, planetary rover attracted a lot of interest and, especially. geological observation brings ouy new knowledge about the history of the birth of planets and our solar system. To get the geological information, the sampling and measurement of a rock and sand are required. It means the measurement of 3D surface information of workspace for manipulator is the most important process. Usually, the surface information is brought out by a stereo image processing, and a camera is the strongest candidate of sensors installing on a rover. However a natural object has much complexity and self-similarity which cause for matching error of stereo process. In this paper, to handle those subjects, we apply the projection of shadow by a stick or a manipulator. Shadow makes featured area helping for matching process and its shape is containing the 3D surface information. Therefore, Shadow Casting Stereo Imaging (SCSI) and Shadow Range Finder (SRF) are proposed and applied for 3D surface measurement. SCSI is based on stereo processing which has occlusion problem, and SRF take monocular image. We discuss data composition with SCSI and SRF to eliminate uncertainty and occlusion area of SCSI. Finally, proposed method is evaluated by experimental setup with manipulator.