基于NSCT的遥感图像模糊增强算法  

Remote Sensing Image Fuzzy Enhancement Algorithm Based on NSCT

作　　者：杜超本[1] 贾振红[1] 覃锡忠[1] 杨杰[2] 胡英杰 李殿均[1] 

机构地区：[1]新疆大学信息科学与工程学院,乌鲁木齐830046 [2]上海交通大学图像处理与模式识别研究所,上海200240 [3]奥克兰理工大学知识工程与发现研究所,新西兰奥克兰1020

出　　处：《计算机工程》2012年第4期188-190,共3页Computer Engineering

基　　金：科技部国际科技合作基金资助项目(2009DFA12870);教育部促进与美大地区科研合作与高层次人才培养基金资助项目(2010-1595)

摘　　要：针对传统小波变换增强方法无法调整图像亮度的问题,提出一种基于非下采样Contourlet变换(NSCT)的遥感图像模糊增强算法。对遥感图像进行NSCT变换,得到图像的高通子带和低通子带,在高通子带设置阈值,对大于阈值的高通子带系数进行线性增强,小于阈值的系数置为0,对低通子带进行模糊对比增强。实验结果表明,该算法能获得较好的图像平均值和熵值,视觉效果较优。Traditional wavelet transformation enhancement can not adjust the brightness of an image effectively. Aiming at this problem, this paper proposes a remote sensing image fuzzy enhancement algorithm based on Nonsubsampled Contourlet Transform（NSCT）. The remote sensing images are transformed into the high-pass sub-band and low-pass sub-band by NSCT, and the threshold is set in the high-pass sub-band, the high-pass subband coefficients which is greater than the threshold is enhanced by linear enhancement, the high-pass subband coefficients which less than the threshold is set zero, the low-pass subband is enhanced by the fuzzy contrast enhancement. Experimental results show the effectiveness of the method in the improvement of entropy and the average.

关 键 词：遥感图像 图像增强 非下采样CONTOURLET变换 模糊增强 自适应阈值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]