基于SOFM的高速图像检索算法实现

广西师范大学学报（自然科学版） ›› 2011, Vol. 29 ›› Issue (2): 180-184.

基于SOFM的高速图像检索算法实现

罗辛¹, 潘乔¹, 王洪亚¹, 陈美², 北研二²

1.东华大学计算机科学与技术学院,上海201620;
2.德岛大学工学院,日本德岛770-8506

收稿日期:2011-05-10 发布日期:2018-11-19
通讯作者: 罗辛(1972—),男,湖南洞口人,东华大学讲师,博士。E-mail:xluo@dhu.edu.cn
基金资助:
国家自然科学基金资助项目(60903160);中央高校基本科研业务费专项资金(11D11203);日本学术振兴会科学研究补助金((B)21300036)

Realization of High-speed Image Search Based on SOFM

LUO Xin¹, PAN Qiao¹, WANG Hong-ya¹, CHEN Mei², KITA Kenji²

1.School of Computer Science and Technology,Donghua University,Shanghai 201620,China;
2.Faculty of Engineering,University of Tokushima,Tokushima 770-8506,Japan

Received:2011-05-10 Published:2018-11-19

摘要/Abstract

摘要： 如何在高维图像数据库中快速有效地检索相似图像一直是图像检索领域的主要研究课题之一。本文提出一种基于自组织特征映射网络(SOFM)的高速图像检索算法,在保留高维空间距离的前提下将图像特征映射到一维空间,在低维空间的限定范围内完成检索工作。通过Corel图像数据库(51 138幅图像)进行的实验表明该算法既明显地提高了检索的速度,又有效地保证了检索的准确率。

关键词: 基于内容的图像检索, 自组织特征映射, 高维特征, 最近邻搜索

Abstract: How to retrieve similar images quickly and effectively from a High-dimensional image database is a fundamental research in theimage retrieval field.A high-speed image retrieval algorithm is proposed basedon self-organizing feature map (SOFM).Using SOFM,high-dimensional vector dataare firstly transformed into one-dimensional units while preserving the higherorder topology by mapping similar data items to the same or the neighboring unit.Experiments on Core image database (51 138 images) show that the proposed algorithm can retrieve images from an image databases,effectively.

Key words: CBIR, SOFM, high-dimensional feature, nearest neighborsearch

中图分类号:

TP391.3

罗辛, 潘乔, 王洪亚, 陈美, 北研二. 基于SOFM的高速图像检索算法实现[J]. 广西师范大学学报（自然科学版）, 2011, 29(2): 180-184.

LUO Xin, PAN Qiao, WANG Hong-ya, CHEN Mei, KITA Kenji. Realization of High-speed Image Search Based on SOFM[J]. Journal of Guangxi Normal University(Natural Science Edition), 2011, 29(2): 180-184.

参考文献

[1] 罗辛,邰晓英,北研二,等.一种基于度量距离学习的图像检索方法[J].广西师范大学学报:自然科学版,2007,25(2):186-189.
[2] 蔡捷飞,袁华.一种基于关键维的图像索引方法[J].广西师范大学学报:自然科学版,2009,27(1):209-212.
[3] BOHM C,BERCHTOLD S,KEIM D.Searching in high-dimensional spaces-index structures for improving the performance of multi-media databases[J].ACMComputing Surveys,2001,33(3):322-373.
[4] 徐勇,杨健,赵英男,等.一种缩减图像维数的方法及其在人脸图像上的应用[J].电子与信息学报,2008,30(1):180-184.
[5] WEBER R,SCHEK H,BLOTT S.A quantitative analysis and performance study for similarity-search methods in high-dimensional spaces[C]//24th International Conference on Very Large Data Bases.New York:Morgan Kaufma-nn,1998:194-205.
[6] KOHONEN T.Self-organizing maps[M].3rd ed.New York::Springer,2001:1-64.
[7] GIONIS A,INDYK P,MOTWANI R.Similarity search inhigh dimensions via hashing[C]//Proceedings of the 25th International Conference on Very Large Data Bases.San Francisco:Morgan Kaufmann,1999:518-529.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed