| 引用本文: | 王小鹏,魏统艺,房超,朱生阳.自适应非局部空间约束与K-L信息的模糊C-均值噪声图像分割算法[J].控制理论与应用,2022,39(7):1261~1271.[点击复制] |
| WANG Xiao-peng,WEI Tong-yi,FANG Chao,ZHU Sheng-yang.Fuzzy C-means algorithm with adaptive non-local spatial constraints and K-L information for noisy image segmentation[J].Control Theory and Technology,2022,39(7):1261~1271.[点击复制] |
|
| 自适应非局部空间约束与K-L信息的模糊C-均值噪声图像分割算法 |
| Fuzzy C-means algorithm with adaptive non-local spatial constraints and K-L information for noisy image segmentation |
| 摘要点击 1507 全文点击 639 投稿时间：2021-10-18 修订日期：2022-06-23 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2022.10989 |
| 2022,39(7):1261-1271 |
| 中文关键词 图像分割 模糊C-均值聚类 非局部空间信息 自适应匹配函数 局部方差绝对差 |
| 英文关键词 image segmentation fuzzy C-means clustering non-local spatial information adaptive matching function local variance absolute difference |
| 基金项目 国家自然科学基金项目(61761027), 甘肃省科技计划项目(20YF8GA036), 甘肃省优秀研究生“创新之星”项目(2021CXZX–610)资助 |
|
| 中文摘要 |
| 针对传统模糊C-均值聚类(FCM)算法对噪声鲁棒性差的问题, 提出一种自适应非局部空间约束与K-L信息
的模糊C-均值噪声图像分割算法. 首先, 通过定义平滑度, 设计自适应匹配函数, 实现非局部空间信息项搜索窗口和
邻域窗口的自适应计算, 克服非局部空间信息窗口大小固定的问题. 其次, 将K-L信息引入目标函数, 利用隐马尔可
夫模型计算图像像素的上下文信息, 减少分割的模糊性. 最后, 利用原始图像和非局部空间信息项局部方差的绝对
差和其倒数自适应约束原始图像和非局部空间信息项, 实现约束项参数的自适应选择, 提高算法的灵活性. 含噪合
成图像和彩色图像分割实验表明, 该算法在分割精准度、平均交互比、归一化互信息、模糊分割系数和模糊划分熵
等性能方面均优于其他几种FCM算法. 例如, 在混合噪声密度为15%的条件下, 算法的模糊分割系数和模糊划分熵
分别达到99.92%和0.14%. |
| 英文摘要 |
| For the conventional fuzzy C-means clustering (FCM) algorithm with poor robustness to noise, a fuzzy Cmeans
noisy image segmentation algorithm with adaptive non-local spatial constraints and K-L information was proposed.
Firstly, by defining the smoothness and designing the adaptive matching function, the adaptive calculation of the search
window and the neighborhood window of the non-local spatial information items was achieved, and the problem of the
fixed size of the non-local spatial information window was overcome. Secondly, the K-L information was introduced into
the objective function, and the hidden Markov model was utilized to calculate the contextual information of image pixels
to reduce the fuzziness of segmentation. Finally, the absolute difference between the local variance of the original image
and the non-local spatial information term and its inverse are employed to adaptively constrain the original image and
the non-local spatial information term, so as to realize the adaptive selection of the parameters of the constraint term and
improve the flexibility of the algorithm. Experiments on noisy synthetic images and color images segmentation show that
the algorithm outperforms several other FCM algorithms in terms of segmentation accuracy, mean intersection over union,
normalized mutual information, fuzzy partition coefficient, and fuzzy partition entropy. For example, the fuzzy partition
coefficient and fuzzy partition entropy of the algorithm reach 99.92% and 0.14%, respectively, under the condition that the
mixed noise density is 15%. |
|
|
|
|
|