Research on Algorithm of Human Gait Recognition Based on Sparse Representation

When the gait feature is recognized, it will be hard to meet the real-time need due to the complexity and long calculated time of the recognition algorithm. Under this circumstance, a novel sparse representation based human gait recognition method is put forth which is based on. First of all, human silhouette is established and gait period is calculated. Next, we adopt Shifting Energy Image (SEI) as the feature of image and then extract Gabor Wavelet and Local Binary Pattern features. Finally, gait feature will be classified and recognized by using sparse representation. CASIA B gait database will be used in the experiment with a view of 90 degrees. The result witnesses that this method has higher recognition rate and can meet the needs of real-time.