Disjoint camera track matching based on an illumination-tolerant major colour spectrum representation algorithm

摘要

Tracking people across disjoint camera views is a challenging task since the appearance of single individuals may vary significantly across such views. A "cumulative colour histogram transformation" algorithm, which is capable of coping with the typical illumination changes occurring, is proposed in this paper. Our disjoint camera track matching is based on a "Moving Object Incremental Major Colour Spectrum Histogram Representation" and post-matching integration algorithms, in which moving objects appearance representation is based on an "online K-means colour-clustering" algorithm. Experimental results have shown that the matching ability of moving objects in two tracks have been improved significantly.