Recursive estimation of absolute motion and structure from two image sequences without stereo correspondence

Abstract

This paper deals with structure and motion estimation by using image sequences captured by multi-camera system moving in a scene. Single camera systems with a relatively small field of view have limited accuracy because of inherent ambiguities of confusion between translation and rotation. The cameras constituting multi-camera system considered ultimately in this paper are arranged so that there is no or small intersection of fields of view. This configuration has advantages for improving the accuracy of the structure and motion estimation because the ambiguities mentioned above decrease due to a large field of view. We propose a recursive algorithm for real time applications of multi-camera system using long term image sequences, which have appearing and disappearing point features. We show in experiments that using just two cameras can obtain much better accuracy than using single camera. Some peculiar implementation techniques contribute to a fast convergence rate and high stability in spite of large initial uncertainty. The method combining stereo and motion is a useful vision cue to estimate both the motion of the cameras and the structure of a scene with an absolute scale factor. Most of the works related to the method have used both motion correspondence and stereo correspondence. Because it is information redundancy to use these two correspondences simultaneously, they can give accurate results. However, the methods to solve the stereo correspondence problem are not yet reliable enough to be used in practical applications, while the motion correspondence problem is easier to solve because sequences of images can be taken at short time intervals. Therefore, if an absolute scale factor can be acquired by only using motion correspondence without stereo correspondence, it is not necessary to use stereo correspondence for accuracy at the cost of reliability. This paper also shows that using a multi-camera system of milt-viewpoint structure has the additional...