基于非局部均值的单帧图像超分辨率重建算法

摘要

In order to obtain a high-resolution image from a single-frame low-resolution image ,a super-resolution reconstruction algorithm based on non-local means is proposed .T he image super-resolution reconstruction is regarded as an inverse problem .A regularization model is established .Non-local means filtering is added with the local structure information and similarity redundant at different scales of natural images fully considered .T he experimental results show that the proposed algorithm can not only effectively inhibit image edge contour ,image details ,and artificial artifacts ,but also naturally robust to noise .%为了解决单帧低分辨率图像获得高分辨率图像的问题，提出了一种基于非局部均值滤波的单帧图像超分辨率算法，将图像超分辨率重建视为反问题，建立正则化模型，充分考虑图像的局部结构信息和自然图像中不同尺度的相似性冗余，加入非局部滤波。实验结果表明，文中算法从单帧图像重建的图像边缘轮廓和纹理较传统算法清楚，有效抑制了人工伪影，同时对噪声具有鲁棒性。在视觉效果及峰值信噪比上都取得良好的结果。