遥感图像信号不相关随机噪声去除方法研究  

Research on signal-independent random noise elimination method for remote sensing images

作　　者：刘雪峰[1] 王聪聪 张现军[1] LIU Xuefeng;WANG Congcong;ZHANG Xianjun(School of Automation ＆ Electronic Engineering,Qingdao University of Science ＆ Teelmology,Qingdao 266000,Chin)

机构地区：[1]青岛科技大学自动化与电子工程学院,山东青岛266000

出　　处：《现代电子技术》2018年第14期62-65,69,共5页Modern Electronics Technique

基　　金：国家自然科学基金(61401244);国家自然科学基金(61773227);山东省高等学校科技计划项目(J15LN39);山东省科技发展计划项目(2013YD01033)~~

摘　　要：遥感图像无论在获取还是在传输中都会受到噪声的干扰,影响图像质量和进一步的数据挖掘。因此,在一般数字图像去噪算法的基础上探讨了对遥感图像信号不相关随机噪声的去除方法。由于小波变换理论在图像去噪处理中的应用广泛且效果显著,在此着重研究基于小波变换的小波阈值去噪算法。进而在coif3小波函数下进行仿真实验,结合仿真结果比较硬阈值和软阈值对噪声滤除的效果,同时还在软阈值的基础上尝试了多级软阈值去噪算法。实验结果表明,软阈值处理后的去噪效果要优于硬阈值,而多级软阈值处理后的效果优于纯软阈值处理的效果。The remote sensing image can be disturbed by noises during its acquisition or transmission,which affects the image quality and further data mining. Therefore,a signal-independent random noise elimination method for remote sensing images is discussed on the basis of the general digital image denoising algorithm. As the wavelet transform theory is widely applied in image denoising processing and has a significant effect,the wavelet threshold denoising algorithm based on wavelet transform is emphatically researched. The simulation experiment was carried out with the coif3 wavelet function. In combination with the simulation results,the noise-filtering effects of hard threshold and soft threshold were compared,and the multi-level software threshold denoising algorithm was tried on the basis of the software threshold. The experimental results show that the denoising effect of soft threshold processing is better than that of hard threshold processing,and the effect of the multi-level soft threshold processing is better than that of the pure soft threshold processing.

关 键 词：图像去噪 遥感图像 信号不相关随机噪声 小波变换 小波阈值 去噪效果 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]