结合空间语义信息的图像表示方法  

Combining Spatial and Semantic Information to Represent Image

作　　者：赵悦[1] 于剑[1] 谢博鋆[1] 

机构地区：[1]北京交通大学计算机与信息技术学院,北京100044

出　　处：《计算机科学与探索》2013年第10期896-904,共9页Journal of Frontiers of Computer Science and Technology

基　　金：国家自然科学基金重点项目No.61033013;高等学校博士学科点专项科研基金No.20120009110006~~

摘　　要：近年来词袋(bag-of-words,BoW)模型因为其较高的性能而被人们认可。词袋模型的改进方法主要包括两种,一种是在图像特征表示中加入空间信息,另一种是加入语义信息。研究了结合图像特征点间的空间和语义信息的高性能图像特征表示方法,通过计算图像中视觉词间的分布距离,并提取相似的视觉词组成视觉短语,来更好地表示图像。在UIUC-Sports8图像库和Scene-15图像库上进行图像分类实验,并与传统的词袋模型及其他模型进行比较,结果显示视觉词短语方法获得了更高的分类准确率。Recently, bag-of-words （BOW） model has been There are mainly two categories of bag-of-words models. approved by researchers due to their good performance. One is to add spatial information into the image feature representation and the other is to add semantic information. This paper proposes an image feature representation method which combines the spatial information between feature points with the semantic information, and makes the feature show better performance. This paper extracts similar visual words by computing distribution divergence and forms visual phrase, which can present the meaning of image. Image classification experiments based on this method are conducted on UIUC-Sports8 dataset and Scene-15 dataset, and the results show that the visual phrase method has better classification accuracy compared with the conventional bag-of-words model and other models.

关 键 词：特征表示 词袋模型 视觉词 视觉短语 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]