An accurate active shape model for facial feature extraction

Abstract

The active shape model (ASM) has been used successfully to extract the facial features of a face image under frontal view. However, its performance degrades when the face concerned is under perspective variations. In this paper, a modified shape model is proposed to make the model represent a face more flexibly, under different orientations. The model of the eyes, nose and mouth, and the face contour are separated. An energy function is defined that links up these two representations of a human face. Three models are employed to represent the important facial features under different poses. A genetic algorithm (GA) is applied to search for the best representation of face images. Experiments show a better face representation under different perspective variations and facial expressions than the conventional ASM can.