摘要
K-means聚类算法在图像分割领域中的运用越来越普遍,但由于K-means算法对噪声具有敏感性,对初始聚类中心具有依赖性,并且容易收敛至局部最优解,使其在图像分割时效果并不是很理想,对此提出一种改进的结合动态粒子群优化与K-means聚类的混合算法来优化图像分割的效果。首先利用双边滤波进行平滑降噪处理,再通过动态调整惯性系数来提高PSO算法的全局优化能力,随后将动态粒子群优化的输出结果作为K-means算法的初始聚类中心,最后通过多次迭代直至收敛。实验结果表明,新算法能有效提升图像分割效果与分割质量。
K-means clustering algorithm is used more and more widely in the field of image segmentation, but because K-means algorithm is sensitive to noise, dependent on the initial clustering center, and easy to converge to the local optimal solution, the effect of K-means clustering algorithm in image segmentation is not very ideal. Proposes an improved combination of dynamic particle swarm optimization. The hybrid algorithm of K-means clustering and K-means clustering is adopted to optimize image segmentation effect. Firstly, bilateral filtering is applied for smoothing noise reduction, and then the global optimization ability of PSO algorithm is improved by dynamically adjusting the inertia coefficient. Then, the output of dynamic particle swarm optimization is used as the initial clustering center of K-means algorithm. Finally, iterations are carried out to converge. Experimental results show that the proposed algorithm can effectively improve the image segmentation effect and quality.
作者
杨雨航
YANG Yu-hang(College of Computer and Information Science,Chongqing Normal University,Chongqing 401331)
出处
《现代计算机》
2019年第8期63-67,共5页
Modern Computer
关键词
图像分割
平滑滤波
粒子群优化
K-MEANS聚类
Image Segmentation
Smoothing Filtering
Particle Swarm Optimization
K-means Clustering
