摘要
针对遥感图像的模糊问题,设计一种基于非下采样剪切波变换与稀疏先验的图像复原算法.首先,利用遥感图像在非下采样剪切波分解下的高频图像的稀疏特性设置先验条件构造图像复原模型;其次,采用交替方向乘子法求解模型;再次,采用软阈值方法对高频图像进行约束处理,在低频图像进行导向滤波处理,以最大可能保留图像的细节信息;最后,将高频图像与低频图像进行重构,对重构后的图像采用卷积神经网络进行深度去噪,最终复原出清晰的图像.将该去模糊算法与H-PNP,GSR,L2TV算法进行实验对比.实验结果表明,该算法能有效去除遥感图像中的模糊和噪声,保留图像的边缘细节,客观评价指标均高于其他3种对比实验算法.
Aiming at the blurring problem of remote sensing images,we designed an image restoration algorithm based on non-subsampled shearlet transformation and sparse prior.Firstly,the image recovery model was created by setting the sparse a priori condition of remote sensing image under non-subsampled shearlet decomposition of the high--frequency image.Secondly,the model was solved by using the alternating direction multiplier method.Thirdly,the high--frequency image was restricted by the soft thresholding method,and the guided filtering was conducted in the low-frequency image to maintain the detaile d information of the image as much as possible.Finally,the high--frequency image and the low-frequency image were reconstructed,the reconstructed image was subjected to deep denoising by using convolutional neural networks,ultimately restoring a clear image.The deblurring algorithm was compared with H-PNP,GSR,and L2TV algorithms through experiments.The experimental results show that the algorithm can effectively remove blurring and noise in remote sensing images,preserve the edge details of the image,and the objective evaluation indexes are higher than the other three comparative experimental algorithms.
作者
成丽波
董伦
李喆
贾小宁
CHENG Libo;DONG Lun;LI Zhe;JIA Xiaoning(School of Mathematics and Statistics,Changchun University of Science and Technology,Changchun 130022,China)
出处
《吉林大学学报（理学版）》
CAS
北大核心
2024年第1期106-115,共10页
Journal of Jilin University:Science Edition
基金
国家自然科学基金(批准号:12171054)
吉林省教育厅科学技术研究项目(批准号:JJKH20230788KJ)。
关键词
遥感图像
非下采样剪切波变换
稀疏先验
图像去模糊
交替方向乘子法
remote sensing image
non-subsampled shearlet transformation
sparse prior
image deblurring
alternating direction multiplier method
