西安电子科技大学学报

• 研究论文 • 上一篇    下一篇

利用核技巧和迭代量化的图像检索算法

杨源1;查宇飞2;覃兵2;梁兵兵3;杨丽薇4;李运强2   

  1. (1. 空军工程大学 空管领航学院, 陕西 西安 710051;
    2. 空军工程大学 航空航天工程学院, 陕西 西安 710038;
    3. 空军西安飞行学院, 陕西 西安 710300;
    4. 空军工程大学 科研部, 陕西 西安 710051)
  • 收稿日期:2015-11-11 出版日期:2017-02-20 发布日期:2017-04-01
  • 作者简介:杨源(1981-),男,博士,E-mail: yangyuankgd@126.com
  • 基金资助:

    国家自然科学基金资助项目(61202339, 61472442, 61203268);航空科学基金资助项目(20131996013)

Image retrieval based on kernel trick and iterative quantization

YANG Yuan1;ZHA Yufei2;QIN Bing2;LIANG Bingbing3;YANG Liwei4;LI Yunqiang2   

  1. (1. College of ATC Navigation, Air Force Engineering Univ., Xi'an 710051, China;
    2. College of Aeronautics and Astronautics Engineering, Air Force Engineering Univ., Xi'an 710038, China;
    3. Air Force Xi'an Flight Academy, Xi'an 710300, China;
    4. Science Research, Air Force Engineering Univ., Xi'an 710051, China)
  • Received:2015-11-11 Online:2017-02-20 Published:2017-04-01

摘要:

针对图像检索过程中数据出现线性不可分和准确度低的问题,提出了一种利用核技巧和迭代量化得到哈希码的图像检索算法.该算法利用核技巧,将图像数据巧妙地从低维映射到高维,从而基本实现数据的线性可分,保证了训练得到哈希码的有效性; 在训练哈希函数的过程中,采用迭代量化的方式,使得图像数据逼近哈希码; 最终实现量化误差的最小化,并生成哈希码以便进行图像检索.在两种图像库上进行测试.实验结果表明,文中算法的检索效果相对于比较的哈希算法得到了一定的提升.

关键词: 图像检索, 哈希, 核技巧, 迭代量化, 线性可分

Abstract:

This paper proposes a novel algorithm for solving the problem of data linear inseparable and low-accuracy in the image retrieval field. In order to get hash codes, the algorithm takes account of kernel trick and iterative quantization. First, the kernel trick is used to map the image data from low-dimension into high-dimension cleverly. In this way the data become linearly separable, and the trained hash codes are proved to be effective. Second, in the process of training the hash function, iterative quantization is used to quantize the image data to the closest hash codes. Finally, the quantitative error is minimized, and the hash codes are generated for image retrieval. Experimental results show that it certainly outperforms other compared hashing algorithms on two image benchmarks.

Key words: image retrieval, hashing, kernel trick, iterative quantization, linear separable