Development Of Human Facial Expression Recognition System

Abstract

Two approaches are proposed for face detection. These approaches are used for expression recognition. In our first approach, the face detection is done using the correlation among face and hair region. The skin color tones are used to locate the region of a face. Further, the hair portion is decided by the process of color spectrum of the hairs. In the second approach, the combination of three color space (RGB, YCbCr and HSI) is used to detect the face. Further, the proposed approach finds the facial features such as eye and mouth regions. Using Rule-based technique, the facial expressions are classified into normal, sad, surprise and smile.

Keywords: Face Detection, Color Model, Polynomial Model, Features Extraction, Facial Expression Recognition