计算机与现代化 ›› 2010, Vol. 1 ›› Issue (5): 50-52.doi: 10.3969/j.issn.1006-2475.2010.05.015

• 图像处理 • 上一篇    下一篇

一种基于LS-SVR滤波器的图像放大方法的研究

关 燕,于忠党   

  1. 渤海大学信息科学与工程学院,辽宁 锦州 121000
  • 收稿日期:2010-01-11 修回日期:1900-01-01 出版日期:2010-05-10 发布日期:2010-05-10

Study of Image Magnification Method Based on LS-SVR Filter

GUAN Yan, YU Zhong-dang   

  1. College of Information Science and Engineering, Bohai University, Jinzhou 121000, China
  • Received:2010-01-11 Revised:1900-01-01 Online:2010-05-10 Published:2010-05-10

摘要: 针对传统插值法的缺点提出了本文的方法。首先介绍了最小二乘支持向量机的原理以及线性平滑滤波,然后构造了LS-SVR滤波算子。根据滤波策略和LS-SVR的特点构造的LS-SVR卷积模板,将LS-SVR的训练过程简化为简单的加权求和运算,提高了算法的运算速度也简化了运算过程的复杂性。

关键词: 图像放大, LS-SVR滤波器, 滤波算子, 卷积模板

Abstract: The paper is presented based on the deficiency of traditional interpolation. First, the theory of LS-SVR and linearity smoothness filtration are introduced, then LS-SVR filtration operator is constructed. To improve the arithmetic operation speed and simplify the arithmetic process complexity, it constructs the LS-SVR convolution mask based on the filtering strategy and LS-SVR characteristic, training process of LS-SVR is turned into simple weighted summation operation.

Key words: image magnification, least square support vector regression (LS-SVR) filter, filtration operator, convolution mask

中图分类号: