摘要
结合人眼视觉特性,设计了一组多尺度、多方向Gabor滤波器,引进粗糙集中的互信息熵,以评价经不同滤波器提取的特征对聚类的影响,去除影响聚类的不重要、冗余因素,突出了区分能力强的特征,使纹理图象的分割质量和算法效率均得到了提高。
This paper designed a set of multi-scale,multi-directional real Gabor filters Incorporating into the human visual characteristics, the concept of mutual information entropy in rough set was introduced to evaluate the effect of the features extracted from different filters on clustering, get rid of the redundancy features and gave prominence to the features which had good distinguishing abilities. Experimental results indicated that the proposed algorithm outperforms conventional approaches in terms of both objective measurements and visual evaluation.
出处
《武汉理工大学学报》
EI
CAS
CSCD
北大核心
2007年第5期134-137,共4页
Journal of Wuhan University of Technology
基金
国家自然基金(60572048)
关键词
纹理分割
GABOR滤波器
粗糙集
信息熵
texture segmentation
Gabor filters
rough set
information entropy