摘要
针对模糊局部信息C-均值(fuzzy local information C-means,FLICM)聚类算法因其局部空间信息的局限性而导致图像分割结果存在误差的问题,改进FLICM算法的相似度测量因子,并考虑邻域空间距离、灰度信息以及灰度方差对分割效果的影响,提出一种用于图像分割的模糊局部信息C均值的修正算法(WFLICM).实验结果表明,WFLICM能够估算邻域像素的衰减程度,提高图像的分割性能,在抑制噪声的同时更好地保留图像细节,且具有更好的抗噪鲁棒性.
FLICM（fuzzy local information C-means）fails to resolve the misclassification problem due to the limitation of local spatial information.In order to solve this problem,a modified FLICM is proposed for image segmentation,which improves the similarity measurement factor by taking into account the effects of spatial distance information,gray level and variance of gray level of neighborhood pixels.The modified algorithm（WFLICM）can accurately estimate the damping extent of neighboring pixels and can suppress noise at large scale while preserving more image details.Experimental results show that the algorithm can improve the performance of image segmentation and has better robustness to noise.
出处
《兰州交通大学学报》
CAS
2016年第1期25-29,共5页
Journal of Lanzhou Jiaotong University
基金
国家自然科学基金(61461025)
甘肃省自然科学基金(148RJYA011)
关键词
模糊C均值
聚类
图像分割
邻域信息
灰度信息
灰度方差
fuzzy C-means
clustering
image segmentation
neighborhood information
gray level information
variance of gray level
