Robust Measurement of Human Faces

摘要

Assessing and modeling human faces are important tasks in many applications, such as computer animation, medical simulation, and face recognition. In this paper, we proposed to reconstruct three-dimensional (3-D) geometry of human faces from projection images of specially designed color encoded pattern. An automatic algorithm was designed for robustly decoding and matching pattern elements. We also tried to increase the spatial resolution by a n a ccompanying corner de tection algorithm Experimental results have shown that the 3-D recovered shape mimics reality and the face model is reliably reconstructed.