A Snake Model Based on Improved Internal Energy

Through traditional Snake model, contour points could not be converged accurately towards object boundary of deeply concave region and complex region. Aiming at this shortcoming, this paper improved internal energy from two aspects: convergent force from contour points to their center and area surrounded by contour points. And the paper used greedy algorithm of alternation of improved Snake model and traditional Snake model. Results show that initial contour points are converged object boundary accurately and quickly in segmentation of complex image, without disturbance of noise, and adjustment of parameters is easy.