Fast place recognition with plane-based maps

摘要: This paper presents a new method for recognizing places in indoor environments based on the extraction of planar regions from range data provided by hand-held RGB-D sensor. We propose to build plane-based map (PbMap) consisting set 3D patches described simple geometric features (normal vector, centroid, area, etc.). world representation is organized as graph where nodes represent and edges connect planes that are close by. structure permits efficiently select subgraphs representing local neighborhood observed planes, will be compared with other corresponding neighborhoods acquired previously. To find candidate match between two we employ an interpretation tree working partially missing planes. The candidates further checked out rigid registration test, which also gives us relative pose matched places. experimental results indicate proposed approach efficient way solve this problem, satisfactorily even when there substantial changes scene (lifelong maps).