摘要
去除噪声与保持图像细节特征是含噪声图像分割中面临的一对矛盾。为此,提出一种改进的模糊C均值算法,通过引入非局部加权距离以抑制噪声影响。其中,权值通过局部图像块距离的指数形式计算,并利用半局部统计特性自适应调整其光滑参数。实验结果表明,新方法具有较强的抗噪声能力,同时能够保持较多地细节特征。
It is a difficult task to wipe out noises and keep more fine information simultaneously for FCM and its variants.In this paper,a modified fuzzy c-means method based on non-local weighted distance is presented.The non-local weighted distance is a linearly-weighted sum distance and the patch difference is used to compute the weight which measures the affinity of two pixels.In the computation of the weights,a local smooth parameter was used adaptive to the semi-local statistics.Validation studies were performed on the synthetic and real-world images with various noises,as well as MR brain images.Experiments results show that the proposed method is very robust to noise and can keep more fine structures.
出处
《模糊系统与数学》
CSCD
北大核心
2011年第3期154-162,共9页
Fuzzy Systems and Mathematics
基金
国家自然科学基金资助项目(61072118)
关键词
图像分割
模糊C均值
非局部加权距离
图像结构信息
Image Segmentation
Fuzzy c-means
Non-local Weighted Distance
Image Structure Information
