改进遗传算法优化模糊均值聚类中心的图像分割

针对传统模糊均值聚类算法存在的问题,提出一种改进遗传算法优化模糊均值聚类中心的图像分割算法.首先在标准遗传算法的交叉操作中引入方向因子,使参与交叉的个体向最佳个体靠近,加快算法的收敛速度,并通过增强群体间的信息共享机制提高算法的全局搜索能力,避免了早熟收敛,改善了全局解的精度;然后采用改进遗传算法选择模糊均值聚类算法的初始聚类中心,实现图像分割;最后采用仿真实验测试算法性能.实验结果表明,相对于传统模糊均值聚类算法及其他图像分割算法,本文算法在分割正确率、分割速度及鲁棒性上均更优.

Image Segmentation Based on Improved Genetic AlgorithmOptimizing Fuzzy Means Clustering Center

In order to improve the image segmentation accuracy, in view of the problems in the traditional fuzzy clustering algorithm, the author proposed an image segmentation algorithm based on improved genetic algorithm optimizing fuzzy means clustering center. First of all, the direction factor was introduced into the crossover operation of standard genetic algorithm to make individual in cross approach to the best individual so as to accelerate the convergence speed, and inter group information sharing mechanism was enhanced to improve the algorithm’s global search capability and avoid the premature convergence so as to improve the accuracy of global solution. Then the initial cluster centers of fuzzy k means clustering algorithm were selected by improved genetic algorithm to realize image segmentation. Finally the performance was tested by simulation experiments. The experimental results show that compared with the traditional fuzzy C means clustering algorithm and other images segmentation algorithm, the proposed algorithm is better in segmentation accuracy rate, the segmentation speed and robustness.