Efficient decoupled pose estimation from a set of points

This paper deals with pose estimation using an iterative scheme. We show that using adequate visual information, pose estimation can be performed in a decoupling the estimation of translation and rotation. More precisely, we show that pose estimation can be achieved iteratively as a function of only three independent unknowns, which are the translation parameters. An invariant to rotational motion is used to estimate the camera position. Once the camera position is estimated, we show that the rotation can be estimated efficiently using a direct method. The proposed approach is compared against two classical methods from the literature. The results show that using our method, pose tracking in image sequences and the convergence rate for randomly generated poses are improved.