|
|||||
|
|
| 基于Vague融合的自动图像标注方法 | |
| 引用本文： | 虎晓红,钱旭,兰洋.基于Vague融合的自动图像标注方法[J].郑州大学学报(理学版),2010,42(1). |
| 作者姓名： | 虎晓红 钱旭 兰洋 |
| 作者单位： | 1. 河南农业大学,信息管理学院,河南,郑州,450002;中国矿业大学,机电与信息工程学院,北京,100083 2. 中国矿业大学,机电与信息工程学院,北京,100083 3. 信阳师范学院,计算机系,河南,信阳,464000 |
| 基金项目： | 教育部科研基金重点资助项目,编号107021 |
| 摘 要： | 提出了一种基于Vague融合的自动图像标注方法,通过有效区域匹配方式,利用近邻语义信息来平衡正负样本数目,并且首次利用Vague集的真假隶属度融合图像的区域信息,从而获得更准确的标注结果.实验结果表明,该标注方法是可行的,同时,与传统的标注方法相比,标注结果得到了明显的提高. |
| 关 键 词： | 信息融合 Vague集 自动图像标注 |
A Vague Fusion Approach to Automatic Image Annotation |
|
| HU Xiao-hong , QIAN Xu , LAN Yang.A Vague Fusion Approach to Automatic Image Annotation[J].Journal of Zhengzhou University:Natural Science Edition,2010,42(1). | |
| Authors: | HU Xiao-hong QIAN Xu LAN Yang |
| Institution: | HU Xiao-hong1,2,QIAN Xu2,LAN Yang3(1.School of Information , Management Science,Henan Agricultural University,Zhengzhou 450002,China,2.School of Mechanical Electronic , Information Engineering,China University of Mining , Technology,Beijing 100083,3.School of Computer , Information Technology,Xinyang Normal University,Xinyang 464000,China) |
| Abstract: | A novel Vague fusion approach to automatic image annotation is presented,which balances the positive and negative labeled images with the semantic information of its neighbors using the adaptive geometric relationship between instances.Then the information of image regions is fused by the true and false members of Vague set.The experiments over Corel images show that this approach is effective for image annotation. |
| Keywords: | information fusion Vague set automatic image annotation |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |