Detection and Filtering of Landmark Occlusions using Terrain Spatiograms

Damian M. Lyons, Fordham University


A team of robots cooperating to quickly produce a map needs to share landmark information between team members so that the local maps can be accurately merged. However, a landmark visible to one robot may be partially occluded to another! Terrain Spatiograms are a landmark representation in which the image spatial information relates to the scene rather than the image. This makes it possible to identify and filter potential landmark occlusions. We present an approach to identifying and filtering occlusions using Terrain Spatiograms, and we report experimental results on 20 landmark datasets for varying states of occlusion. We show that occlusion can be detected and filtered, resulting in improved landmark matching scores.