The paper presents an artificial fish swarm algorithm based on K-Means clustering.The algorithm uses the feature of having artificial fish swarm algorithm's strong robustness and being not easy to fall into local optimum value,and hence dynamically determines the number of clusters and center,overcoming the defects of K-Means clustering initial point selected unstable.The image segmentation is processed based on the fusion of two algorithms.The test proves the algorithm is ideal.%提出一种基于K-Means聚类的人工鱼群算法,该算法利用人工鱼群算法鲁棒性较强且不易陷入局部最优值的特点,动态的确定了聚类的数目和中心,解决了K-Means聚类初始点选择不稳定的缺陷,在此两种算法融合的基础上进行图像分割处理,经试验证明该算法效果理想.
展开▼