Date Added: Jun 2009
This paper investigates the design of a system for recognizing objects in 3D point clouds of urban environments. The system is decomposed into four steps: locating, segmenting, characterizing, and classifying clusters of 3D points. Specifically, the authors first cluster nearby points to form a set of potential object locations (with hierarchical clustering). Then, they segment points near those locations into foreground and background sets (with a graph-cut algorithm). Next, they build a feature vector for each point cluster (based on both its shape and its context).