A Novel Kernel-Based Fuzzy C-Means Algorithm with Spatial Information for Image Segmentation
Fuzzy c-means (FCM) algorithm is a popular method widely used in the image segmentation field. However, the conventional FCM algorithm is sensitive to noise because of not utilizing the spatial information in the image. Meanwhile, it is also unsuitable for revealing non-Euclidean structure of the input data due to the use of Euclidean distance. To overcome the drawbacks, this paper proposes a novel kernel-based FCM algorithm that incorporates the spatial information into the membership function for clustering (NSKFCM). In the algorithm, we replace the original Euclidean distance with a Gaussian kernel-induced distance, then use the distribution statistics of the neighborhood pixels and the weight coefficient based on the distance attributes to form a new membership function. Experimental results show that NSKFCM can segment images more effectively and provide more robust segmentation results.
image segmentation fuzzy c-means kernel function spatial information
Yue Yang Shuxu Guo Runlan Tian Peng Liu
College of Electronic Science and Engineering, Jilin University, Changchun, China College of Electronic Science and Engineering, Jilin University, Changchun, China Aviation Electroni
国际会议
成都
英文
732-736
2010-12-17（万方平台首次上网日期，不代表论文的发表时间）