Undecimated Wavelet Transform based classification of human emotion

In this paper, an approach for human emotion recognition system based on Undecimated Wavelet Transform (UWT) is presented. The main drawback of Discrete Wavelet Transform (DWT) is not translation invariant. Translations of an image lead to different wavelet coefficients. UWT is used to overcome this and more comprehensive feature of the decomposed image is obtained. The classification of human emotional state is achieved by extracting the energies from all sub-bands of UWT. The robust K-Nearest Neighbor (K-NN) is constructed for classification. The evaluation of the system is carried on using JApanese Female Facial Expression (JAFFE) database. Experimental results show that the proposed UWT based human emotion recognition system produces more accurate recognition rate than DWT. For 3^rd level decomposition, UWT based features produces 82% classification rate while DWT based features produces 73.22%. The maximum classification rate achieved by the proposed system is 85.62% using 5^th level decomposition.