基于概念格层次分析的视觉词典生成方法.pdfVIP

第27 卷 第 1 期 计算机辅助设计与图形学学报 Vol. 27 No.1 2015 年 1 月 Journal of Computer-Aided Design Computer Graphics Jan. 2015 基于概念格层次分析的视觉词典生成方法 * 钟利华, 张素兰, 胡立华, 张继福 (太原科技大学计算机科学与技术学院 太原 030024) (zhsulan@126.com) 摘 要: 视觉词典容量是影响图像场景分类精度的重要因素之一, 大容量的视觉词典因计算量较大影响了分类的效 率, 而小容量的视觉词典由于多义词问题的严重致使场景分类精度降低. 针对该问题, 提出一种基于概念格层次分 析的视觉词典生成方法. 首先生成关于训练图像视觉词包模型的初始视觉词典; 然后在构造的概念格上利用概念格 的概念层次性, 通过动态地调整外延数阈值, 获取粒度大小不同容量的描述图像各场景语义的约简视觉词典; 最后 对各类约简视觉单词构成向量进行异或, 删除多义词, 进而生成有效描述图像场景语义的视觉词典. 实验结果表明, 文中方法是有效的. 关键词：视觉词包; 视觉词典; 多义词; 概念格; 层次分析 中图法分类号：TP391.4 A Concept Lattice Hierarchy Based Generating Method of Visual Dictionary * Zhong Lihua, Zhang Sulan , Hu Lihua, and Zhang Jifu (School of Computer Science and Technology, Taiyuan University of Science and Technology, Taiyuan 030024) Abstract: The visual dictionary size is an important factor that affects the performance of scene classifica- tion. The large capacity of visual dictionary can influence the classification efficiency due to the lager cal- culation, while the small capacity of visual dictionary can reduce the classification accuracy because of the influences of polysemy. To solve the problem, a generating method of visual dictionary based on the concept lattice hierarchy is proposed in this paper. First, the initial visual dictionary of training images on bag-of-visterms model is generated. Then, with the use of concept lattice’s hierarchy analysis, the different granularities of reduced visual dictionaries are extracted from the concept lattice by setting different exten- sion thresholds. Finally, the polysemy is deleted by making XOR operations on all types of the reduced vis- ual dictionarie