摘要
从明暗恢复形状(SFS)是计算机视觉中三维形状恢复问题的关键技术之一。本文从Hamilton系统理论角度解释了SFS问题的动力学本质,简要介绍了求解相应Hamilton-Jacobi方程的三种经典方法,即特征线方程法、粘稠解方法以及等值线求解方法,并分析了各方法的优缺点。
Shape from Shading (SFS) is one of crucial techniques in shape recovery in computer vision. This paper uncovers the dynamical essence of shape from shading problem from the Hamilton system point of view. At the same time three classical methods, namely characteristic strips equations, viscosity solutions, and level sets, are introduced to solve the corresponding Hamilton-Jacobin equation of SFS problem. And their advantages and disadvantages are discussed respectively.
出处
《系统仿真学报》
CAS
CSCD
2001年第S2期30-32,共3页
Journal of System Simulation
