Decomposition of Range Images using Markov Random Fields (TA-L1)

Author(s) : Andreas Pichler (Profactor Produktionsforschung Gmbh., Austria) Robert B. Fisher (School of Informatics, University of Edinburgh, UK) Markus Vincze (Automation and Control Institute, Austria)

Abstract : This paper describes a computational model for deriving a decomposition of objects from laser rangefinder data. A general decomposition rule is based on concave discontinuities and local shape characteristic to produce the constituent parts of an object. The knowledge about the process is encoded in a Markov Random Field (MRF). Shape index and curvedness descriptors along with discontinuity and concavity distributions are introduced to classify region labels correctly. In addition, a novel way to classify the shape of a surface is proposed resulting in a better distinction of concave, convex and saddle shapes. To achieve a reliable classification of shape classes a multi-scale method provides a stable estimation of the shape index. Experimental results are presented to demonstrate the proposed approach.

Menu