An Image Segmentation Algorithm Based on the Simulated Annealing and Improved Snake Model
Liqun Tang, Kejun Wang, Guangsheng Feng, Yonghua Li · 2007
For segmenting the regions in the image exactly and quickly. The segmentation algorithm based on the simulated annealing and the improved snake model was proposed. The improved snake model is that centre-energy is added on the traditional snake model and curve energy can be tuned according to the position of the dot. Therefore the initialized curve not only keeps the topological properties, but also can fit the concave of the object. Firstly, the edge can be roughly fitted according to the improved snake model. Secondly, the edge is exactly fitted based on the SA. The optimal result can be gotten in term of the SA. It is proved by the experiments that the segmentation result is very valid.