作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2007, Vol. 33 ›› Issue (24): 189-190. doi: 10.3969/j.issn.1000-3428.2007.24.066

• 人工智能及识别技术 • 上一篇    下一篇

基于序列子集共轭梯度的超分辨率图像重构

吴 强1,刘 琚2,乔建苹2,王行愚3   

  1. 1. 驻511厂军代表室,南京 210002;2. 山东大学信息科学与工程学院,济南 250100; 3. 华东理工大学信息科学与工程学院,上海 200037
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-12-20 发布日期:2007-12-20

Super-resolution Image Reconstruction Based on Sequence of Subset Conjugate Graduate Optimization

WU Qiang1, LIU Ju2, QIAO Jian-ping2, WANG Xing-yu3   

  1. 1. The No.511 Factory of the PLA, Nanjing 210002; 2. School of Information Science and Engineering, Shandong University, Jinan 250100; 3. School of Information Science and Engineering, East China University of Science & Technology, Shanghai 200037
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-12-20 Published:2007-12-20

摘要: 提出一种基于序列子集共轭梯度最优化算法的超分辨率图像重建算法。通过图像配准算法估计得到不同低分辨率图像间的平移和旋转量,结合期望图像的统计先验对问题进行规整,建立优化的代价函数。利用序列子集共轭梯度最优化迭代算法求解,得到高分辨率图像。仿真结果表明,该算法可以使内存降低15%,运算速度提高20%。

关键词: 图像重构, 超分辨率, 共轭梯度

Abstract: A sequence of subset conjugate graduate optimization based super-resolution algorithm is proposed for super-resolution, where image registration algorithm is used to estimate the displacement and rotation between different low-resolution images, prior knowledge is used for regulation and building the cost function, and the high-resolution image is estimated by using sequence subset conjugate graduate optimization algorithm. Simulation results demonstrate that the method can realize the super-resolution with 15% memory reduction and 20% high speed promotion.

Key words: image reconstruction, super-resolution, conjugate graduate

中图分类号: