Human Emotions Recognition Based On Wavelet Transform

In this paper, the authors proposes facial expression recognition algorithm of human being based on wavelet transform. The input image is decomposed using discrete wavelet transform (DWT) into different frequency sub-bands. Normally facial emotions are recognized by changes in facial features like changes in eyebrows, eye and mouth etc. But as per human age changes, it is difficult to recognize emotion. So we adopt new approach based on wavelet transform which is little bit different than the other techniques. The objective is to recognize facial expression using energy components of image and efficiently use wavelet transform for calculating energy components of image. The proposed algorithm recognized different human emotions efficiently and energy of spectral component is utilized well