New generalized GVF external forces for snakes

Snakes, or active contours, are used extensively in computer vision and image processing applications, particularly to locate object boundaries. This paper describes the formulation of active contours and develops their external force: Potential and Non Potential forces. A Non Potential external force for active contours, called gradient Vector Flow (GVF) was developed to address problems belonging to initialization and poor convergence to boundaries concavities. GVF is computed as a diffusion of the gradient vectors of a gray level or binary edge map derived from the image. Therefore, the corresponding snake is formulated directly from the condition of forces balance rather than a variationnal formulation. In this paper, the GVF is introduced and the GVF is generalized including the spatially varying weighting functions. Two Generalized GVF fields are defined. This is expected to improve active contour convergence to long, thin boundary indentations, while maintaining other desirable properties of GVF, such as extended capture range.