A Learning Algorithm for Place Recognition

We present a place recognition algorithm for SLAM systems using stereo cameras that considers both appearance and geometric information. Both near and far scene points provide information for the recognition process. Hypotheses about loop closings are generated using a fast appearance technique based on the bag-of-words (BoW) method. Loop closing candidates are evaluated in the context of recent images in the sequence. In cases where similarity is not sufficiently clear, loop closing verification is carried out using a method based on Conditional Random Fields (CRFs). We compare our system with the state of the art using visual indoor and outdoor data from the RAWSEEDS project, and a multisession outdoor dataset obtained at the MIT campus. Our system achieves higher recall (less false negatives) for full precision (no false positives), as compared with the state of the art. It is also more robust to changes in appearance of places because of changes in illumination (different shadow configurations in different days or time of day). We discuss the promise of learning algorithms such as ours, where learning can be modified on-line to re-evaluate the knowledge that the system has about a changing environment.