Three-dimensional human head modeling is useful in video-conferencing or ot
her virtual reality applications. However, manual construction of 3D models
using CAD tools is often expensive and time-consuming. Here we present a r
obust and efficient method for the construction of a 3D human head model fr
om perspective images viewing from different angles. In our system, a gener
ic head model is first used, then three images of the head are required to
adjust the deformable contours on the generic model to make it closer to th
e target head. Our contributions are as follows. Our system uses perspectiv
e images that are more realistic than orthographic projection approximation
s used in earlier works. Also, for shaping and positioning face organs, we
present a method for estimating the camera focal length and the 3D coordina
tes of facial landmarks when the camera transformation is known. We also pr
ovide an alternative for the 3D coordinates estimation using epipolar geome
try when the extrinsic parameters are absent. Our experiments demonstrate t
hat our approach produces good and realistic results. (C) 2001 Academic Pre
ss.