计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (15): 200-201.DOI: 10.3778/j.issn.1002-8331.2009.15.058

• 图形、图像、模式识别 • 上一篇    下一篇

融合区域特征和相关反馈的图像检索

苏光伟1,张敏情1,杨晓元1,2   

  1. 1.武警工程学院 电子技术系网络与信息安全武警部队重点实验室,西安 710086
    2.西安电子科技大学 ISN国家重点实验室,西安 710071
  • 收稿日期:2008-03-25 修回日期:2008-06-16 出版日期:2009-05-21 发布日期:2009-05-21
  • 通讯作者: 苏光伟

Image retrieval based on region features and relevance feedback

SU Guang-wei1,ZHANG Min-qing1,YANG Xiao-yuan1,2   

  1. 1.Network and Information Security Key Lab,Dept. of Electronics,Eng. College of the Armed Police Force,Xi’an 710086,China
    2.State Key Lab of Integrated Service Networks,Xidian University,Xi’an 710071,China
  • Received:2008-03-25 Revised:2008-06-16 Online:2009-05-21 Published:2009-05-21
  • Contact: SU Guang-wei

摘要: 提出了一种新的基于内容的图像检索算法,该算法提取图像的颜色-空间特征,在HSV空间中将图像按照H分量进行区域划分,利用区域特征层面上的相似度对图像进行检索,并引入用户的相关反馈来调整并记录示例图像中各对象的特征权值。实验结果表明,该算法可以使计算机更加精确地理解用户的查询要求,提高查询的准确率。

Abstract: The retrieval algorithm uses a method based on color-spatial features,translates the images into HSV color space,and cuts the images into regions with the H values.The algorithm uses the relevance feedback from the users into the retrieval algorithm,and proposes a method to change and save the information users provided during the retrieval progress.The experiment shows that the algorithm can greatly improve the retrieval precision.