基于局部最大和最小强度先验的遥感图像盲去模糊  

Blind Deblurring of Remote Sensing Images Based on Local Maximum and Minimum Intensity Priors

作　　者：王琪瑶 胡琸悦 李潇雁 陈凡胜[1,2] Wang Qiyao;Hu Zhuoyue;Li Xiaoyan;Chen Fansheng(Key Laboratory of Intelligent Infrared Perception,Shanghai Institute of Technical Physics,Chinese Academy of Sciences,Shanghai 200083,China;Hangzhou Institute for Advanced Study,National University of Defense Technology,Zhejiang 310024,Hangzhou,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院上海技术物理研究所中国科学院智能红外感知重点实验室,上海200083 [2]国防科技大学杭州高等研究院,浙江杭州310024 [3]中国科学院大学,北京100049

出　　处：《激光与光电子学进展》2023年第4期419-426,共8页Laser & Optoelectronics Progress

摘　　要：针对遥感图像的运动模糊问题,提出一种基于局部最大和最小强度先验的遥感图像盲去模糊方法。该方法利用遥感图像局部像素强度的稀疏性作为先验条件,使用简单的迭代阈值收缩方法求解潜像和模糊核,再由非盲反卷积算法得到去模糊图像。实验结果表明,所提方法能提高计算效率,对于可见、红外遥感图像,均能有效恢复图像的纹理细节,抑制伪影,提升了复原图像的主观效果与客观评价指标。A blind deblurring method of remote sensing images based on local maximum and minimum intensity priors is proposed to solve the motion blur problem.The sparsity of local pixel intensity of remote sensing image is used as a prior condition in this method,and a simple iterative threshold shrinkage method is applied to solve the latent image and blur kernel,then we obtain the deblurred image using by nonblind deconvolution algorithm.The experimental results show that the proposed method can improve the computational efficiency.For both optical and nearinfrared remote sensing images,it can availably restore the texture details of the images,suppress artifacts,and improve the subjective effect and objective evaluation index for the restored images.

关 键 词：遥感图像 运动模糊 图像去模糊 盲复原 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]