"computer vision; omni-directional images; stereo-matching; stereographic projection"@en . "21230" . . . "2"^^ . "R" . . . "[675523E60DD8]" . "Stereo Matching Insensitive to the Change of Camera Orientation" . . "We study the geometry of the projection with the aim to facilitate stereo-matching of images acquired by the NIKON fisheye 180 deg. field-of-view lens camera. For the stereo matching, we assume that our target objects are locally planar. First, we formulate the homography for spherical cameras since all central omni-directional cameras can be converged to a spherical camera model. Second, we analyze the mapping between two omni-directional images through numerical experiments. Then, we consider the most suitable image representation of omni-directional images to obtain stereo-correspondences." . . "Stereo Matching Insensitive to the Change of Camera Orientation"@cs . . . . "Stereo Matching Insensitive to the Change of Camera Orientation"@en . "501607" . "We study the geometry of the projection with the aim to facilitate stereo-matching of images acquired by the NIKON fisheye 180 deg. field-of-view lens camera. For the stereo matching, we assume that our target objects are locally planar. First, we formulate the homography for spherical cameras since all central omni-directional cameras can be converged to a spherical camera model. Second, we analyze the mapping between two omni-directional images through numerical experiments. Then, we consider the most suitable image representation of omni-directional images to obtain stereo-correspondences."@en . "Stereo Matching Insensitive to the Change of Camera Orientation" . . "Stereo Matching Insensitive to the Change of Camera Orientation"@cs . . "We study the geometry of the projection with the aim to facilitate stereo-matching of images acquired by the NIKON fisheye 180 deg. field-of-view lens camera. For the stereo matching, we assume that our target objects are locally planar. First, we formulate the homography for spherical cameras since all central omni-directional cameras can be converged to a spherical camera model. Second, we analyze the mapping between two omni-directional images through numerical experiments. Then, we consider the most suitable image representation of omni-directional images to obtain stereo-correspondences."@cs . "Stereo Matching Insensitive to the Change of Camera Orientation"@en . . "RIV/68407700:21230/06:03124669!RIV09-MSM-21230___" . . "2"^^ . . . "Pajdla, Tom\u00E1\u0161" . "Torii, Akihiko" . . "RIV/68407700:21230/06:03124669" .