基于新的遗传算法的模糊C均值聚类用于遥感图像分割

标准FCM对噪声十分敏感,并且依赖于初始聚类中心选择,算法通常得到的是局部最优解而非全局最优解。针对此问题提出一种基于猴王遗传算法的改进的FCM算法.猴王遗传算法是一种新颖的全局优化搜索算法,具有高效的计算性能和优良的全局搜索能力。本文首次将猴王遗传算法（MKGA）与结合空间领域信息的FCM相结合,利用改进的FCM算法的目标函数建立适应度函数,利用猴王遗传算法搜索全局最优解,代替FCM的基于梯度下降的迭代过程,从而有效地避免了模糊C-均值聚类算法收敛到局部最优和对噪声敏感的问题。在此基础上实现了对遥感图像的聚类分割。实验结果表明,该算法对于遥感图像显示了较好的分割效果和较强的抗噪能力。

A new FCM algorithm based on monkey-king genetic algorithm for remote sensing image segmengtation

FCM clustering algorithm has been proven effective for image segmentation.However,the conventionally standard FCM algorithm is sensitive to noise and it is dependent on the choice of the initial distribution of cluster center,so consequently the algorithm ends up in a local optimum.Monkey-King Genetic（MKGA） is a novel global optimization search algorithm genetic algorithm.This paper proposes a new segmentation algorithm;which combines thc Monkey-King Genetic Algorithm and FCM.Then the fitness function contained neighbor information is set up according to the object function in FCM algorithm.By applying Monkey-King genetic algorithm,we can achieve the global optimum.The proposed method can effectively avoid getting into the local optimum solution and more robust to noise.The experiments on the segmentation results demonstrates that the algorithm performs the effective ability of searching global optimal solution and more robust to noise than the standard FCM algorithm and GA-FCM algorithm.