序列图像超分辨重建技术研究

Research on the sequence image super-resolution reconstruction

利用静止轨道光学相机完成对地高分辨率观测,对国计民生具有重大意义。然而,通过增大光学相机口径、增长相机焦距、减小像元尺寸等技术手段提升图像分辨率不可避免地会带来一些其他的问题,如制造难度加大、成本上升、信噪比降低。因此,基于数字图像处理技术,结合动量更新策略,提出了一种基于最大后验概率估计的超分辨重建算法。该算法通过对多帧具有一定像素偏移的图像进行处理,最终实现了分辨率的提升。此外,针对超分辨重建在实际应用场景中没有一个非常好的评价指标问题,将无参考图像的像质评价算法BRISQUE引入超分辨重建中。仿真和实验均表明,所提出的算法相比以往算法取得了更好的效果,对于图像噪声、模糊均具有较好的鲁棒性,可以有效地实现图像超分辨重构,提升图像的视觉效果。

The use of geostationary orbit optical camera to complete high-resolution observation of the earth has great significance to national economy and people′s livelihood.However,increasing the resolution of the image by increasing the caliber of the optical camera,increasing the focal length of the camera,reducing the pixel size,inevitably bring about some other problems.Such as manufacturing more difficult,rising costs,lower signal to noise ratio.We proposes a super-resolution reconstruction algorithm through the method of maximum posterior probability estimation,which is based on digital image processing technology and the momentum update strategy.To improve the resolution of image,our algorithm processes multi-frame images with different pixel offsets.In addition,as there is no good index to evaluate a superresolution reconstruction algorithm in practical application scenes,it proposes applying the non-reference image quality evaluation algorithm BRISQUE（blind reference-less image spatial quality evaluators）to super-resolution reconstruction.Both simulations and experiments show that the algorithm proposed in this paper has better performances than previous ones.Besides,it has better robustness for the problems of noise and blurring of the image and it can effectively reconstruct super-resolution images and obviously improve the visual effects of images.