联合L_(1)和L_(0)先验模型的超分辨率重建算法  

Super-resolution reconstruction algorithm by combining L_(1) and L_(0) prior models

作　　者：李利 尹增山 石神 Li L;YIN Zengshan;SHI Shen(Innovation Academy for Microsatellites,Chinese Academy of Sciences,Shanghai 201203,China;University of Chinese Academy of Sciences,Beijing 100049,China;School of Information Science and Technology,ShanghaiTech University,Shanghai 201210,China)

机构地区：[1]中国科学院微小卫星创新研究院,上海201203 [2]中国科学院大学,北京100049 [3]上海科技大学信息科学与技术学院,上海201210

出　　处：《中国科学院大学学报（中英文）》2022年第3期369-376,共8页Journal of University of Chinese Academy of Sciences

基　　金：科技部国家重点研发计划项目(2017YFB0502902)资助。

摘　　要：超分辨率重建可以从低分辨率图像序列中重建出高分辨率图像,提高图像质量。重建出边缘保持且噪声低的高分辨率图像,仍具有挑战。针对此问题,在L_(1)先验模型中添加图像梯度的L_(0)范数作为先验知识,提出联合L_(1)和L_(0)先验模型的超分辨率重建算法,既保留L_(1)先验模型边缘保持的优点,又保留L_(0)先验模型抑制噪声的优点。将该算法与双三次插值、Total Variation(TV)先验模型和L_(1)先验模型作对比,通过仿真实验数据和真实实验数据的分析,验证本文算法的有效性。Super-resolution(SR)reconstruction can reconstruct a high-resolution image from low-resolution image sequences and improve image quality.Reconstructing a high-resolution image with edge preserving and low noise is still a challenge in SR.Therefore,the L_(0) norm of the image gradient is added as prior knowledge in the L_(1) prior model,and a SR reconstruction algorithm by combining the L_(1) and L_(0) prior model is proposed in this paper,which not only retains the advantage of L_(1) prior model preserving edges,but also retains the advantage of L_(0) prior model suppressing noise.Compared with bicubic interpolation,total variation(TV)prior model,and L_(1) prior model,the validity of the algorithm is verified through the analysis of simulation experimental data and real experimental data.

关 键 词：超分辨率重建 L_(1)先验模型 L_(0)先验模型 噪声抑制 双三次插值 

分 类 号：TN911.73[电子电信—通信与信息系统]