Novel snake model with wavelet-based energy minimization

The external force of original snake model by integrating gradient directional information explicitly and then derive a new active contour model. Thus, in minimizing the snake energy, the proposed new model employs both the gradient magnitude and gradient direction information, which can provide better segmentation results. Based on dynamic programming approach to snake model, a wavelet-based multiresolution algorithm is proposed in this article for energy minimization. Combined with Mallat's wavelet-based edge detection theory, the proposed multiresolution algorithm can greatly decrease the amount of computation, become more flexible in selecting filter than continuation method, and has less initial position sensitivity and parameter sensitivity which makes the setting of parameters simplified.