Moving part image partitioning for walking gait identification

摘要

Described herein is an apparatus and method for gait recognition. The apparatus includes circuitry that is configured to receive a gait sequence including a predetermined number of image frames of a subject. The received gait sequence is processed to generate a gait-energy-image (GEI). A plurality of Gabor filter responses is computed by convoluting the generated GEI with a bank of Gabor filters, wherein the filter bank includes a first predetermined number of unique scales, and a second predetermined number of unique orientations. Further, the circuitry is configured to partition, each Gabor filter response of the computed plurality of Gabor filter responses, into a predetermined number of overlapping regions and extract, a predetermined number of statistical features only from the overlapping regions, the extracted statistical features corresponding to texture consent of the subject. The circuitry eventually recognizes the subject based on a classification of the extracted statistical features.