Gait-Based Recognition of Human Using an Embedded Hidden Markov Models

An embedded hidden Markov models (e-HMM) gait recognition scheme based on gait energy image (GEI) is proposed. First, the mean GEI is calculated from gait periodic, then we analyze the mean GEI regions, making use of the two dimensional discrete cosine transform (2D-DCT) to transfer the regions into observation vector, and complete the e-HMM training and humans recognition. We compare the proposed algorithm with other gait recognition approaches on USF HumanID Database and CASIA Gait Database. Experimental results show that the proposed approach is valid and has encouraging recognition performance.