Human action recognition using silhouette histogram

This paper presents a method for human action recognition using silhouette histogram. The human silhouette is obtained by using the background subtraction method and then mapped into three different polar coordinate systems that characterize three parts of a human figure respectively. Each polar coordinate system is quantized by partitioning it into several cells with different radii and angles. The evaluation applied at the Weizmann dataset shows that our method can achieve better performance than the existing silhouette-based system.