Person re-identification based on set-to-set metric learning

People re-identification is a difficult problem in non-overlapping video surveillance, because pedestrian images contain variations in view angle, lighting, background clutter and occlusion. This paper presents an approach for person re-identification in surveillance system by taking pedestrian sequence (set) as processing element. The distance metric is learnt from relative and irrelative set-to-set pairs by extending relative distance comparison. Combining the appearance features, the person re-identification is transformed as a set-to-set matching based on the learnt distance. The approach proposed in this paper is compared with several state-of-the-art algorithms on NLPR_MCT dataset. Experimental results show that this algorithm has a more superior performance than the state-of-the-art algorithms.