Estimation of the center of rotation and 3D motion parameters from stereo sequence images and virtual validation using three-COMERO
Wanpeng Cao,Rensheng Che,Dong Ye
DOI: https://doi.org/10.1016/j.patrec.2007.03.020
IF: 4.757
2007-01-01
Pattern Recognition Letters
Abstract:In this paper, we present new a kind of motion model to estimate an object's center of rotation and 3D motion parameters based on stereo sequence images, and validate the algorithm through virtual simulations based on three-COMERO. We assume that an accurate match between the correspondence points has been accomplished successfully in advance. First, we set up two virtual space reference frames on the motion object and the left camera, respectively, and figure out their space relation through the coordinates of the correspondence points that lie on the measured motion object. Secondly, we design the object's 3D motion model based on the object's rigid motion character under the two reference frames. From the object's motion model, dynamics, and knowledge of the rigid motion, we work out the motion algorithm to compute the 3D motion parameters. Thirdly, using these relevant motion parameters, the center of rotation may be calculated through eliminating the effect of the translational motion. Finally, the motion equations are deduced from these motion models, and the closed-form solutions are figured out. Virtual simulation experiments through three-COMERO are conducted, using synthesized stereo sequence images corrupted by noise, and the experimental results prove the model's correctness and the algorithm's validity.