计算机应用研究2011，Vol.28Issue(10)：3986-3988,3.DOI:10.3969/j.issn.1001-3695.2011.10.106
单尺度词袋模型图像分类方法
Single-scale image classification employing Bag-of-Words model
摘要
Abstract
The general image classification methods relying on SIFT feature description need to construct multi-scale space, thus it is not only time-consuming but also irrelevant to visual sense. This paper proposed a new image classification method. It directly extracted single-scale SIFT features based grid,and described the features employing Bag-of-Words( BOW) model afterwards. Because single-scale SIFT need not build multi-scale space and retains more global information,the proposed method could reduce the computational complexity substantially and improved the classification accuracy significantly. Kxperimental results illustrate that compared with the standard SIFT based BOW model,the classification accuracy of BOW model formed from single-scale SIFT is significantly improved.关键词
图像分类/单尺度SsIFT/视觉单词/词袋模型Key words
image classification/ single-scale SIFT/ visual word/ BOW model分类
信息技术与安全科学引用本文复制引用
陈凯,肖国强,潘珍,李正浩..单尺度词袋模型图像分类方法[J].计算机应用研究,2011,28(10):3986-3988,3.基金项目
国家自然科学基金资助项目(61003203) （61003203）
重庆市自然科学基金资助项目(CSTC2010 BB2230) （CSTC2010 BB2230）
