Robust Facial Feature Localization Using Improved Active Shape Model and Gabor Filter

Face location under the environment of illumination chagne is a serious problem, in order to reduce illumination effect, in this paper, we propose a facial feature location method based on improved active shape model (IASM) and Gabor filter to reduce the intensity contrast and to enhance edge information. First, face images are preprocessed by the proposed illumination normalization method using Gabor wavelets. Then, reducing Gabor kernels based on the similarity of Gabor feature vector, the location of facial features can fit more efficient and fast by the proposed feature-based weighted warping. Through our proposed method, it not only obtains the better face alignment, but also overcomes the active shape model (ASM) method which caused the failure for aligned target. Experimental results demonstrate that the proposed system is superior to active shape model (ASM) method for Yale database with varied illumination database (Yale database) or JAFFE database with seven type facial expressions.