Place Recognition using Surface Entropy Features

In this paper, we present an interest point detector and descriptor for 3D point clouds and depth images, coined SURE, and use it for recognizing semantically distinct places in indoor environments. We propose an interest operator that selects distinctive points on surfaces by measuring the variation in surface orientation based on surface normals in the local vicinity of a point. Furthermore, we design a view-poseinvariant descriptor that captures local surface properties and incorporates colored texture information. In experiments, we compare our approach to a state-of-the-art feature detector in depth images (NARF). Our descriptor achieves superior results for matching interest points between images and also requires lower computation time. Finally, we evaluate the use of SURE features for recognizing places.