Region Based Human Gait Identification Using SVM Classifier

Correct identification of person from a distance is an important issue in the field of visual surveillance and monitoring applications. To identify the person while their walking, Gait play an important role. During their walking, every parts of the human body move differently. Which part of the body, contributes more for person identification, on the basis of this, we have developed rectangular region based silhouette analysis (RRSA) algorithm to evaluate the contribution of individual parts of the body to identify the person correctly. This algorithm is tested on CASIA Gait database by using support vector machine (SVM) classifier and wavelet feature extraction method. Experimental result shows that the proposed algorithm is not only fast but also more effective.