期刊信息
- 刊名： 河北师范大学学报（自然科学版）Journal of Hebei Normal University (Natural Science)
- 主办： 河北师范大学
- ISSN： 1000-5854
- CN： 13-1061/N
- 中国科技核心期刊
- 中国期刊方阵入选期刊
- 中国高校优秀科技期刊
- 华北优秀期刊
- 河北省优秀科技期刊
Image Classif ication Based on Multi2level Tolerance Granulation
摘要/Abstract
摘要：
在网络最大流算法的研究中,为了减少计算量,提出了许多改进的方法.基于图论中的最大流最小割定理,利用网络流图的对偶图的最短路径求网络最大流,对求最短路径的Dijkstra算法进行了研究,给出了一种改进的Dijkstra算法模型,该算法采用了堆排序中的小根堆来选择最短路径结点,使用集合运算对堆中的结点进行处理,使得参加运算的结点数减少,提高了算法的效率。
Abstract：
The spatial distribution knowledge and hierarchical information of image features were obtained by using local features of image blocks. An approach to hierarchical classification of images based on tolerance granular space was presented,which showed the principle that image classification was done from coarse to finegrained levels based on granular computing.The experiment results showed that the proposed approach was effective.
关键词
Key words：
granular computing
;
tolerance granular space
;
image partition
;
image feature
;
image classification