This paper presents a novel approach to shape modeling using front propagation and level set methods. The authors, Ravikanth Malladi, James A. Sethian, and Baba C. Vemuri, introduce a technique that can model complex shapes, including those with significant protrusions and sharp corners, without prior assumptions about the object's topology. The method is based on the ideas developed by Osher and Sethian for modeling propagating solid/liquid interfaces with curvature-dependent speeds. The interface, represented as a level set, moves along its gradient field with a speed that depends on the curvature. The speed term is synthesized from image data to stop the interface near object boundaries. The resulting equation of motion is solved using entropy-satisfying upwind finite difference schemes. The paper discusses various computational aspects, including narrow bands, reinitializations, and different stopping criteria. Experimental results on synthetic and low-contrast medical images demonstrate the effectiveness of the method. The authors conclude that their approach offers advantages over existing methods, such as being relatively independent of initial guesses and capable of handling complex topological changes.This paper presents a novel approach to shape modeling using front propagation and level set methods. The authors, Ravikanth Malladi, James A. Sethian, and Baba C. Vemuri, introduce a technique that can model complex shapes, including those with significant protrusions and sharp corners, without prior assumptions about the object's topology. The method is based on the ideas developed by Osher and Sethian for modeling propagating solid/liquid interfaces with curvature-dependent speeds. The interface, represented as a level set, moves along its gradient field with a speed that depends on the curvature. The speed term is synthesized from image data to stop the interface near object boundaries. The resulting equation of motion is solved using entropy-satisfying upwind finite difference schemes. The paper discusses various computational aspects, including narrow bands, reinitializations, and different stopping criteria. Experimental results on synthetic and low-contrast medical images demonstrate the effectiveness of the method. The authors conclude that their approach offers advantages over existing methods, such as being relatively independent of initial guesses and capable of handling complex topological changes.