[1]刘咏梅,代丽洁.词间相关性的CMRM图像标注方法[J].智能系统学报,2011,6(4):350-354.
LIU Yongmei,DAI Lijie.A method of CMRM image annotation based on interword correlation[J].CAAI Transactions on Intelligent Systems,2011,6(4):350-354.
点击复制
LIU Yongmei,DAI Lijie.A method of CMRM image annotation based on interword correlation[J].CAAI Transactions on Intelligent Systems,2011,6(4):350-354.
词间相关性的CMRM图像标注方法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
6
期数:
2011年第4期
页码:
350-354
栏目:
学术论文—人工智能基础
出版日期:
2011-08-25
- Title:
- A method of CMRM image annotation based on interword correlation
- 文章编号:
- 1673-4785(2011)04-0350-05
- Keywords:
- image annotation; CMRM; correlation; wordcorrelation matrix
- 分类号:
- TP391
- 文献标志码:
- A
- 摘要:
- 自动图像标注因其对图像理解和网络图像检索的重要意义，近年来已成为新的热点研究课题.在图像标注的CMRM模型基础上，提出了一种基于词间相关性的CMRM标注方法.该方法提取了标注字之间的词间相关关系，并利用图学习算法，通过将词间相关性矩阵叠加到初始标注矩阵的方法对标注结果进行了改善.利用Corel5k标注图像库中的自然场景图像进行实验.实验结果表明，该方法很好地完成了对测试集图像的自动标注，在查全率与查准率上较CMRM模型有所提高.
- Abstract:
- Automatic image annotation is significant for image understanding and retrieval of web images. As a result, it has become a hot research topic in recent years. On the basis of a CMRM image annotation model, an efficient image annotation method was proposed based on interword correlations. The correlations between the annotated words were picked up, and the interword correlation matrix was added to the initial labeling matrix by a graph learning algorithm. The proposed annotation approach was tested by a Corel5k database with natural scene images. The experimental result shows that the method can automatically label the images in the test set very well. The recall and precision are increased compared to CMRM.
- 参考文献/References:
-
［1］陈世亮，李战怀，袁柳.一种基于区域特征关联的图像语义标注方法［J］.计算机工程与应用， 2007, 43(2): 5357.
CHEN Shiliang, LI Zhanhuai, YUAN Liu. Image semantic annotating method based on region features relevancy［J］. Computer Engineering and Applciations, 2007, 43(2): 5357.
［2］YU Feiyang, HORACE H S．Semantic content analysis and annotation of histological images［J］．Computers in Biology and Medicine, 2008, 38： 635649.
［3］LAVRENKO V, CROFT W. Relevancebased language models［C］//Proceedings of the 24th Annual International ACM SIGIR Conference. New Orleans, USA， 2001： 120127.
［4］LAVRENKO V, CHOQUETTE M, CROFT W. Crosslingual relevance models［C］//Proceedings of the 25th Annual International ACM SIGIR Conference. Tampere, Finland， 2002： 175182.
［5］王斌.图像检索中自动标注与快速相似搜索技术研究［D］.长沙：中国科学技术大学， 2007: 545.
WANG Bin. Researches on automatic image annotation and fast similarity searching technology in CBIR［D］. Changsha： University of Science and Technology of China， 2007: 545.
［6］袁曾任.人工神经元网络及其应用［M］.北京:清华大学出版社, 1999: 4050.
［7］JEON J, LAVRENKO V， MANMATHA R. Automatic image annotation and retrieval using crossmedia relevance models［C］//Proceedings of the 26th International ACM SIGIR Conference. Toronto, Canada, 2003: 119126.
［8］卢汉清，刘静. 基于图学习的自动图像标注［J］.计算机学报, 2008, 31(9): 16291639.
LU Hanqing, LIU Jing. Image annotation based on graph learning［J］. Chinese Journal of Computers, 2008, 31(9): 16291639.
［9］http://www.cs.arizona.edu/people/kobus/research/data/eccv_2002
［10］纪颖. 一种基于实例的图像自动语义标注方法［J］. 哈尔滨理工大学学报, 2009, 14(1): 3842.
JI Ying. An instancebased method for automatic image semantics annotation［J］. Journal of Harbin University of Science and Technology, 2009, 14(1): 3842.
备注/Memo
收稿日期： 2010-03-15.
基金项目：中央高校基本科研业务费专项资金资助项目（HEUCF100604）.
通信作者： 刘咏梅.E-mail:liuyongmei@hrbeu.edu.cn.
作者简介：
刘咏梅，女，1973年生，副教授，硕士生导师，博士. 主要研究方向为模式识别、图像理解和生物信息学.
?代丽洁，女，1983年生，硕士，主要研究方向为图像处理与模式识别、图像标注.
更新日期/Last Update:
2011-09-30