计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (22): 153-155.DOI: 10.3778/j.issn.1002-8331.2009.22.050

• 图形、图像、模式识别 • 上一篇    下一篇

基于NSCT的红外与可见光图像融合

钟庆新,夏利民   

  1. 中南大学 信息科学与工程学院,长沙 410075
  • 收稿日期:2008-04-28 修回日期:2008-10-13 出版日期:2009-08-01 发布日期:2009-08-01
  • 通讯作者: 钟庆新

Non-sampled Contourlet transform based fusion method for infrared and visible images

ZHONG Qing-xin,XIA Li-min   

  1. Department of Information Science and Engineering,Central South University,Changsha 410075,China
  • Received:2008-04-28 Revised:2008-10-13 Online:2009-08-01 Published:2009-08-01
  • Contact: ZHONG Qing-xin

摘要: 提出一种基于非下采样Contourlet变换的红外与可见光图像融合方法。该方法对源图像经非下采样Contourlet变换分解后的高频系数,考虑不同传感器的成像机理进行活性度量,并结合多分辨率系数间相关性来实现加权融合;低频系数则通过一种局部梯度进行活性度量,再采用加权与选择相结合的规则实现融合。最后,通过非下采样Contourlet逆变换重构获得融合图像。实验结果表明了该方法的有效性和可行性。

关键词: 图像融合, 非下采样Contourlet变换, 红外图像, 可见光图像

Abstract: A fusion method for infrared and visible images based on non-sampled contourlet transform is proposed in this paper.After infrared and visible images decomposed through non-sampled contourlet transform in the proposed method,the high frequency coefficients is fused by means of weight averaging method which is via the combination of activity measure using the different imaging mechanism of sensors and the strategy of the relativity of those multi-resolution coefficients while the low frequency coefficients is fused via the combination of weighting and selection through activity measured by a local gradient.Finally,the fused image is reconstructed through inverse non-sampled contourlet transform.The experiment results indicate this fusion method possesses favorable effectivity and feasibility.

Key words: image fusion, non-sampled contourlet transform, infrared image, visible image