Human action recognition employing TD2DPCA and VQ

A novel algorithm for human action recognition in the transform domain is presented. This approach is based on Two-Dimensional Principal Component Analysis (2DPCA) and Vector Quantization (VQ). This technique reduces the computational complexity and the storage requirement by at least a factor of 45.27, and 12 respectively, while achieving the highest recognition accuracy, compared with the most recently published approaches. Experimental results applied on the Weizmann dataset confirm the excellent properties of the proposed algorithm, which lends itself to real-time economic implementation.