3D stereo reconstruction of human faces driven by differential constraints

Conventional stereo algorithms often fail in accurately reconstructing a 3D object because the image data do not provide enough information about the geometry of the object. We propose a way to incorporate a priori informatio n in a reconstruction process from a sequence of calibrated face images. A 3D mesh modeling the face is iteratively deformed in order to minimize an e nergy function. Differential information extracted from the object shape is used to generate an adaptive mesh. We also propose to explicitly incorpora te a priori constraints related to the differential properties of the surfa ce where the image information cannot yield an accurate shape recovery. (C) 2000 Elsevier Science B.V. All rights reserved.