A behavior classification based on Enhanced Gait Energy Image

摘要

A behavior classification method based on Enhanced Gait Energy Image (EGEI) and 2-Directional 2-dimensional principal component analysis ((2D)²PCA) was proposed. EGEI extracted more useful feature information. The high dimensional feature space was reduced to lower dimensional space by (2D)²PCA, which outperformed PCA and 2DPCA.The nearest-neighbor classifier was adopted to distinguish different actions. Experimental results showed that the algorithm was simple, and achieved higher classification accuracy with less running time.