An Optimized Part Based Gait Recognition using Multi-Objective Particle Swarm Optimization

摘要

Gait identification task becomes difficult due to the change of appearance by different cofactors (e.g., shoe, surface, carrying, view, and clothing). Some parts of gait are affected by cofactors and other parts remains unaffected. Most of the gait identification systems consider only most effective parts thereby omitting less effective parts. However some significant features for gait identification resides in less effective parts and are important for more accurate recognition. In this paper, adaptive fusion of part based gait identification is proposed. The proposed gait identification adaptively fuses the best informative less effective part with the most effective parts. The best informative less effective part is selected by using Multi objective adaptive PSO to the varying threshold value. These parts are fused using adaptive fusion method and from these fused parts, the variance ratio is estimated and recognition is done based on variance threshold value. The variance threshold value is calculated based on Particle Swarm optimization (PSO) which dynamically calculates the threshold value for varying parts. Experimental result of proposed system achieves better result when compared with recognition using EnDFT.