Higher order orthogonal moments for invariant facial expression recognition
Digital Signal Processing
| Hi-index | 0.00 |
Automatic facial expression is a challenging task of intelligent human-computer interaction. In this paper, we present a comparative study on moments based feature extraction methods in terms of their capability to recognize facial expression images. The moments include Hu’s moments, Zernike moments, Wavelet moments and Krawtchouk moments. Krawtchouk moments are applied to facial expression analysis for the first time. Experiments are conducted on Cohn-Kanade facial expression database. The test images are original images, localized images and edge detected images. Experimental results show that wavelet moments outperform other moments based methods in facial expression recognition.