| 李昌兴,张晓璐,雷柳.基于粒子群优化非局部模糊聚类图像分割算法[J].光电子激光,2019,30(7):759~767 |
| 基于粒子群优化非局部模糊聚类图像分割算法 |
| Non-local fuzzy clustering image segmentation algorithm based on particle swarm optimization |
| 投稿时间：2019-01-06 |
| DOI： |
| 中文关键词: 非局部均值算法 粒子群优化 模糊因子 |
| 英文关键词:non-local mean algorithm particle swarm optimization the fuzzy factor |
| 基金项目:陕西省自然科学基础研究计划(2018JM1054)资助项目 (1.西安邮电大学 理学院,陕西 西安 710121； 2.西安邮电大学 通信与信息工程学院, 陕西 西安 710121) |
|
| 摘要点击次数: 712 |
| 全文下载次数: 0 |
| 中文摘要: |
| 为提高局部模糊聚类算法(WFLICM)对噪声图像 分割的抗噪性,克服模糊聚类图像分割算法对初 始聚类中心的敏感性及易陷入局部最优问题,在WFLICM算法的基础上提出一种基于粒子群 优化的融合 局部和非局部空间信息的模糊聚类图像分割算法(PSO-WMNLFCM)。首先,利用粒子群优化 算法的全局 寻优能力得到最优粒子,并以此粒子作为模糊聚类算法的初始聚类中心。其次,用像素的非 局部空间信息 替换模糊因子中的局部邻域值,产生新的目标函数。最后,由拉格朗日乘子法最小化目标函 数,得到隶属 度和聚类中心的更新公式,从而完成图像分割。仿真结果表明,PSO-WMNLFCM算法相比于 模糊局部聚 类(FLICM)算法、局部模糊权重(WFLICM)算法、非局部模糊聚类(NLFCM)算法、非局部模 糊聚类 (MNLFCM)算法、基于粒子 群的局部模糊聚类(PSO-FLICM)算法的划分系数提高了20.92%,20.51%,24.84%,1.44%,23.28%左右。 |
| 英文摘要: |
| In order to improve the anti-noise performance of local fuzzy cluster ing algorithm(WFLICM) on noisy image segmentation,and overcome the problem that fuzzy fuzzy clusterin g image segmentation algorithm is sensitive to the initial clustering centers and easy to fall into t he local optimum,on the basis of WFLICM algorithm,we propose a fuzzy clustering image segmentation algorithm based on particle swarm optimization (PSO-WMNLFCM ).This algorithm integrates the non-local spatial information of pixels in the local fuzzy factor and takes the local and non-local spatial info rmation of the image into consideration.Firstly,the global optimization ability of particel swarm optimiz ation algorithm is utilized to obtain the optimal particels,and the particels ar e used as the initial clustering centers of fuzzy clusteri ng algorithm.Secondly,the local neighborhood values of the fuzzy factors are rep laced with non-local spatial information of the pixels to generate a new object ive function.Finally,by minimizing the objective function of Lagrange multiplier method,the update formulas of membership degree and clustering center are obtained,so as to complete image segmentation.Simulation results show that c ompared with fuzzy local clustering (FLICM) algorithm,local fuzzy weeight(WFLICM )algorithm,non-loacal fuzzy clustering(NL FCM)algorithm,non-local fuzzy clustering (MNLFCM) algorithm and local fuzzy clu stering based on particle swarm (PSO-FLICM) algorithm,the partition coefficient of PSO-WMNLFCM algorithm is improved by about 20.92%,20.51%,24.84%,1.44%,23.28%. |
| 查看全文 下载PDF阅读器 |
| 关闭 |
|
|
|
|
|