A FACTORIZATION METHOD FOR FISH-EYE CAMERAS(International Workshop on Advanced Image Technology 2006)

概要

論文の詳細を見る
In this paper, we propose a method for solving the shape reconstruction problem with a fish-eye camera model. The method computes shape and motion from a sequence of omnidirectional images acquired with fish-eye cameras. Each omnidirectional image is transformed into a pure perspective image. Shape and motion are obtained from the sequence of the perspective images by affine iterations. Although in the previous work the camera motion is restricted in a two-dimensional plane, the proposed method has no such restriction about the camera motion. We evaluate the effectiveness of the proposed method in terms of the reconstruction error measured by RMSE. We then show a result of shape reconstruction from real omnidirectional images acquired with a fish-eye camera.
社団法人電子情報通信学会の論文
2006-01-03