Managing Data Association in visual SLAM using SIFT features
This paper describes an approach to solve the Simultaneous Localization and Mapping (SLAM) problem for autonomous mobile robots using visual landmarks and a Rao-Blackwellized particle filter. Our map is represented by a set of three dimensional landmarks referred to a global reference frame. We use significant points extracted from stereo images as natural landmarks. In particular we employ SIFT features found in the environment. Each landmark is associated with a visual descriptor that partially differentiates it from others. We concentrate on a reduced set of highly stable landmarks. In order to do that, we track a visual feature for a significant number of frames prior to integrating it in the filter. As a result, we obtain different examples that represent the same natural landmark. Using this procedure, a better model for each landmark is obtained, which lets us improve data association among the landmarks in the map.