基于AHLO与K均值聚类的图像分割算法

针对图像分割中K均值算法全局搜索能力差、初始聚类中心选择敏感的问题,提出了一种将自适应人类优化算法与K均值算法相结合的聚类算法.该算法利用自适应人类学习优化算法初始化聚类中心,提高K均值算法的稳健性.结果表明,该算法聚类得到的标准差相比传统K均值算法和基于粒子群K均值(PSO-Kmeans)算法分别小两个数量级和一个数量级,同时图像分割得到的PSNR值均较高,具有算法收敛速度更快,聚类质量更好,图像分割效果更好,适应性更强的优点.

Image segmentation algorithm based on AHLO and K-means clustering

Aiming at the problem of poor global searching ability and the sensitivity of initial clustering center selection by K-means algorithm for image segmentation, a clustering algorithm with the combination of adaptive human learning optimization(AHLO)and K-means algorithms was proposed. AHLO algorithm was used to initialize the clustering centers in the proposed algorithm so as to improve the stability of K-means algorithm. The results show that the standard deviation obtained with the proposed clustering algorithm is two orders of magnitude lower than the traditional K-means algorithm and one order of magnitude lower than the PSO-Kmeans algorithm, respectively. Meanwhile, the PSNR values of image segmentation obtained by the proposed algorithm are relatively higher. The as-proposed algorithm has the features of faster convergence speed, better clustering quality, better image segmentation effect and stronger adaptability.