Detection of Human Facial Behavioral Expression Using Image Processing

The paper presents Haar feature extraction classifier based image processing technique for the analysis of human facial behavioral expressions. As an improvisation, some of instances misclassified by Haar classifier are further detected by subsequent classifier adaptive algorithm (Adaboost). Careful attention is paid while generating the templates of various facial behavioral expressions including eye open, close, drowsy and smile. The obtained results create the possibilities of adapting an embedded image processor with the developed algorithms for the real time applications. The image processing algorithm is implemented by using Matlab.