A New Linear Method for Euclidean Motion/Structure from Three Calibrated Affine Views

Long Quan (1) , Yuichi Ohta (2)

1 MOVI - Modeling, localization, recognition and interpretation in computer vision
2 ESYS - College of Engineering Systems [Tsukuba]

We introduce a unified framework for developing matching constraints of multiple affine views and rederive 2-view (affine epipolar geometry) and 3-view (affine image transfer) constraints within this framwork. We then describe a new linear method for Euclidean motion and structure from 3 calibrated affine images, based on insight into the particular structure of these multiple-view constraints. Compared with the existing linear method of Huang and Lee [7], the new method uses different and more appropriate constraints. It has no failure mode of the Euclidean factorisation method of Tomasi and Kanade [18]. We demonstrate the method on real image sequences.

Domaines: Synthèse d'image et réalité virtuelle [cs.GR]

Long Quan, Yuichi Ohta. A New Linear Method for Euclidean Motion/Structure from Three Calibrated Affine Views. IEEE Conference on Computer Vision and Pattern Recognition (CVPR '98), Jun 1998, Santa Barbara, United States. pp.172--178, ⟨10.1109/CVPR.1998.698605⟩. ⟨inria-00590100⟩