Gait recognition stands out as a compelling, recently developed biometric method due to its ability to discreetly identify individuals from afar, even with low-resolution images. Initially, this work uses a gait representation technique, the Gait Energy Image (GEI), to process gait images and generate features. Furthermore, a statistical shape analysis method based on GEI, using corner and centroid extraction, has been developed. This method efficiently captures the statistical shape information via an efficient corner structure descriptor, thus improving gait recognition system performance. Evaluation of the proposed system was performed using the well-known CASIA A and CASIA B gait databases. Results demonstrate that the proposed approach achieves a promising classification accuracy, outperforming some leading contour-based gait recognition techniques.