A measure of happiness and sadness based on facial expression recognition

الملخص

Face recognition systems are one of the most important applications in the field of computer vision. Where these systems enter into the development of many applications, especially in the field of medicine, security and Human Computer Interaction (HCI), In addition to the development of robotics. The system proposed in this paper includes the design of a software model capable of distinguishing two of the most important expressions of the human face: happiness and sadness, plus to the natural expression of the human face. The proposed system includes the approach to the design of any recognition system, starting from image acquisition and pre-processing, through the extraction of attributes. Finally to classify and give the output of the recognition. In the first stage a hybrid algorithm was adopted to extract the oval face. Two dimensions principal component analysis (2DPCA) for stage of features extraction. In the classification stage was used Euclidean distance. The results obtained showed high accuracy to distinguish the proposed expressions, so that the results of the recognition reached (95 %) percent when testing 120 samples from a Multimedia Understanding Group (MUG) Database [1].

https://doi.org/10.24897/acn.64.68.406