结合递归Cycle Spinning的小波-Contourlet变换红外图像增强算法  

Infrared image enhancement algorithm using wavelet-contourlet transform with recursive cycle spinning

作　　者：邓超[1] 侯满义[1] 刘泽乾[1] 王靖华[2] 

机构地区：[1]空军航空大学作战指挥系,吉林长春130022 [2]空军航空大学数学教研室,吉林长春130022

出　　处：《激光与红外》2013年第9期1068-1071,共4页Laser & Infrared

摘　　要：针对Contourlet变换缺乏平移不变性的缺陷,提出了一种基于小波-Contourlet变换的红外图像增强算法,并与递归Cycle Spinning相结合,消除小波-Contourlet变换的失真现象。实验结果表明:与单独使用小波变换和单独使用Contourlet变换相比,本文方法去噪效果更明显,能够获得更高的PSNR值,使图像视觉效果得到改善。For the defect of Contourlet transform that it can not keep shift-invariant,a method for infrared image enhancement based on Wavelet-Contourlet transform is proposed.Recursive Cycle Spinning is also adopted to eliminate the distortion of Wavelet-Contourlet transform.The experiments shows that the proposed method is more outstanding than just using Wavelet transform or Contourlet transform in de-noising.This method can get higher PSNR value and improve the visual effects of infrared image.

关 键 词：小波-CONTOURLET变换 CYCLE SPINNING 平移不变性 图像增强 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TN911.73[自动化与计算机技术—计算机科学与技术]