Model-based active object recognition using MRF matching and sensor planning

摘要

This paper presents an active object reconition algorithm using MAP-MRF matching and sensor planning strategy. The matching between the sensed and model object is based on surface properties. A new measure of surface distinuishability is defined for sensor planning. MAP-MRF framework is used for generating matching label set. A measure of confidence of correcnt match is determined based on the posterior energy. An active object recongition algorithm is used for detrmining the next viewpoint of the camera if ambiguity exists inthe matching result. The mext viewpoint is chosen base the surface with highest distinuishability. Experimental results on images under perfec and imperfect segmentation are presented.