A FEATURE-BASED FACE DETECTOR USING WAVELET FRAMES

In this paper, we propose a novel scheme for detection and precise segmentation of human faces in color images where the number, the location, the orientation and the size of the faces are unknown, under nonconstrained scene conditions such as complex background and uncontrolled illumination. A deformable template is used as a generic model of the face, de ned by stable facial features grouped by anthropometric geometric and textural constraints. The di erent areas of the face template are characterized by extracting simple statistical measures from suitably selected bands of a wavelet decomposition. The candidate face is classi ed by applying a set of optimally ordered heuristic and probabilistic tests on the extracted statistical feature vectors. Experimental results are provided to demonstrate the robustness of our approach and its capability to precisely detect faces under varying scale, expression and orientation.