Estimation of parameter values appearing in space and orientation dependent curve evolution process models

Level sets are a robust and flexible mathematical framework for the representation of evolving curves and surfaces. The heart of the level set approach is a "speed function" that models the physical processes underlying the evolution. Though in general such a model may be quite complex, here we take the point of view that some processes may be characterized reasonably well by a model with relatively few free parameters. Such models would be of significant value for optimal design and real-time control. One possible area of application is thin film manufacturing processes such as etching and deposition. Our goal is to estimate the value of experimental test results. The paper considers specifically the case of speed functions that vary as a function of spatial position and orientation. Examples of such processes are found in lithographic development, and ion-assisted etching. Simple simulations demonstrate the feasibility of the method.