Emotion recognition using facial expression has become an active research topic in recent years. In this paper we present an efficient method for emotion recognition, which has better performance over previous art of works. This work proposes an efficient attempt to investigate the suitability and sensitivity of the thermal imaging technique to detect specific muscles heat patterns and there by predicting the emotions. In this work, feature extraction is carried out by Kernel PCM and emotion classification is performed using Multi Class SVM. Thermal imaging is used for the investigation of Action Unit (AU) productions. A facial AU represents the contraction of a specific muscle or a combination of muscles, and earlier research had demonstrated that such muscle contraction induces an increase in skin temperature. For this reason, thermal imaging analysis might be well suited to detect AU production and there by predicting the emotional state of a person. We used a multi class SVM approach to classify nine different AUs or combinations of AUs and to differentiate their speed and strength of contraction. The Multi class SVM classifier gives promising results for the emotion classification process

1. P. Ekman, W. V. Friesen, and J. C. Hager, Facial Action Coding System. Consulting Psychologist Press, 1978.
2. J. J. J. Lien, T. Kanade, C. C. Li, and J. F. Cohn, "Detection, Tracking and Classification of Action Units in Facial Expression,"IEEE J. Robotics and Autonomous Systems, special issue: face expression in human-robot interaction systems, vol. 31, pp. 131-146, 2000.
3. S. Delplanque, D. Grandjean, C. Chrea, L. Aymard, I. Cayeux, C. Margot, M. I. Velazco, D. Sander, and K. R. Scherer, "Sequential Unfolding of Novelty and Pleasantness Appraisals of Odors: Automatic Reactions, Emotion, vol. 9, no. 3, pp. 316-328, 2009.
4. Z. Zhu, J. Fei, and I. Pavlidis, "Tracking Human Breath in Infrared Imaging," Proc. Fifth IEEE Symp. Bioinformatics and Bioeng. , pp. 227-231, 2005
5. J. Gonza´lez-Alonso, B. Quistorff, P. Krustrup, J. Bangsbo, and B. Saltin, "Heat Production in Human Skeletal Muscle at the Onset of Intense Dynamic Exercise," J. Physiology, vol. 524, pp. 603-615, 2000.
6. D. Lundqvist and J. E. Litton, The Averaged Karolinska Directed Emotional Faces (AKDEF) Dept. of Clinical Neuroscience, Psychology Section, Karolinska Inst. , 1998.
7. Irene Kotsiay, Stefanos Zafeiriouy, Nikolaos Nikolaidisy and Ioannis Pitasy, Multiclass Support Vector Machines and Metric Multidimensional Scaling for Facial Expression Recognition, Aristotle University of Thessaloniki, Department of Informatics Thessaloniki, Greece, 2009
8. Satyanadh Gundimada and Vijayan K. Asari, Facial Recognition Using Multisensor Image Based on Localized Kernel Eigen Spaces, IEEE Transactions On Image Processing, Vol. 18, No. 6, June 2009
9. Sophie Jarlier, Didier Grandjean, Sylvain Delplanque, Karim N'Diaye, Isabelle Cayeux, Maria Ine´sVelazco, David Sander, Patrik Vuilleumier, and Klaus R. Scherer, Thermal Analysis of Facial Muscles Contractions, IEEE transactions on Affective Computing, vol. 2, no. 1, January-March 2011
10. Chung-Hsien Wu, Senior Member, IEEE, and Wei-Bin Liang, Emotion Recognition of Affective Speech Based on Multiple Classifiers Using Acoustic-Prosodic Information and SemanticLabels,IEEE Transactions On Affective Computing,Vol. 2, No. 1, January- March 2011
11. Xin Chen, Patrick J. Flynn, Kevin W. Bowyer, "IR and Visible light face Recognition", University of NotreDame, USA, http://www. identix. com/products/.
12. Bai-Ling Zhang, Haihong Zhang, and Shuzhi Ge, "Face Recognition by Applying Wavelet Subband Representation and Kernel Associative Memory", IEEE Transaction on Neural Networks.

Emotion Recognition Action Unit Thermal Imaging Eigen Faces Kernel Pca