On parameter estimation using level sets

Consider the problem of selecting the member of a parametrized family of curves that best matches a given curve. This is a key step in determining proper values for adjustable parameters in low-order plasma etching and deposition models. Level set methods offer several attractive features for treating such problems. This paper presents a parameter estimation scheme that exploits the level set formulation. The method is completely geometric; there is no need to introduce an arbitrary coordinate system for the curves. Analytic results necessary for the application of gradient descent algorithms are derived, and some preliminary numerical results are presented.