Artificial Fish Swarm Algorithm of Snake Model
LI Shao-fa · Computer Technology and Development · 2010
The result of classical active contour model is related with the initial state and easy to fall into a local minimum.In order to solve these two problems,it can combine active contour model of multi-stage decision-making with fish-swarm decision-making process to put forward the fish active contour algorithm.By the characteristics of active contour model,the way is constructed corresponding to the fish-swarm algorithm,and the active contour model for image segmentation problem is transformed into the optimal path search problem,in order to provide a new approach for obtaining the precise contour.The experimental results show that this method is effective under the initial curve at anywhere in the image.Through mathematical proof and experimental results can be aware of the fish active contour algorithm having good global convergence.It avoids that the original active contour model using the variational method optimization is difficult to reach the global optimum,or even difficult to achieve the local optimum.