北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2011, Vol. 34 ›› Issue (5): 6-9.doi: 10.13190/jbupt.201105.6.wangkp

• 论文 • 上一篇    下一篇

加权特征自动图像标注方法

王科平1,2,王小捷1,钟义信1   

  1. 1.北京邮电大学 计算机学院, 北京 100876; 2.河南理工大学 电气工程与自动化学院, 河南 焦作 454000
  • 收稿日期:2010-10-26 修回日期:2011-05-28 出版日期:2011-10-28 发布日期:2011-08-26
  • 通讯作者: 王科平 E-mail:wangkp@hpu.edu.cn
  • 基金资助:

    国家自然科学基金项目 (90920006); 河南省控制工程重点学科开放实验室开放基金项目(KG200908); 河南省教育厅自然科学基金项目(2011B520017); 河南理工大学青年科学基金项目(Q201133)

A Weighted Feature Based Automatic Image Annotation

  • Received:2010-10-26 Revised:2011-05-28 Online:2011-10-28 Published:2011-08-26
  • Contact: Wang Ke-Ping E-mail:wangkp@hpu.edu.cn

摘要:

提出了一种基于加权特征的图像自动标注方法.该方法首先采用加权特征聚类算法对图像区域进行语义聚类,这种聚类算法根据图像特征的统计分布来计算特征与类别的相关度,增加相关度高的特征的权重,避免聚类算法被弱相关或不相关的特征所支配;然后,根据训练集中样本图像的标注情况建立图像区域与语义关键字的关联;最后,在未标注图像区域给定时,计算每个语义关键字的条件概率,将条件概率最大的语义概念作为图像的标注. 在Corel图像库的数据集上验证了新方法的有效性.

关键词: 加权特征, 图像自动标注, 聚类

Abstract:

An automatic image annotation method based on weighted feature is proposed. Firstly, a weighted feature clustering algorithm is employed on the semantic concept clusters of the image regions. For a given cluster, we determine relevant features based on their statistical distribution and assign greater weights to relevant features as compared to less relevant features.In this way the computing of clustering algorithm can avoid dominated by trivial relevant or irrelevant features. Then, the relationship between clustering regions and semantic concepts is established according to the labeled images in the training set. Under the condition of the new unlabeled image regions, we calculate the conditional probability of each semantic keyword and annotate the new images with maximal conditional probability. Experiments on the Corel image set show the effectiveness of the new algorithm.

Key words: weighted feature, automatic image annotation, clustering

中图分类号: