超模式斜采样遥感图像超分辨复原方法

超模式斜采样成像技术通过将两排探测器阵列倾斜成像提高相机空间分辨率。由于不满足香农采样定理．图像中出现混叠。为了降低混叠对遥感图像超分辨复原的影响，文章提出了一种结合最佳倒易晶胞的超分辨复原方法。该方法根据超模式斜采样频域对偶网格建立相应的混叠模型．通过权重函数衡量噪声和混叠在频谱中的分布，得到噪声、混叠最小最佳倒易晶胞；并基于Huber—MRF（Markov Random Field）图像先验模型建立了耦合最佳倒易晶胞的MAP（Maximuma Posterior）超分辨复原模型。实验表明，该方法降低了超模式斜采样成像中混叠对复原的影响，获得了良好的超分辨复原效果。

Super-Resolution Restoration of Super-tilting Mode Remote Sensing Image

Super-tilting mode improves the spatial resolution by rotating the two dedicated arrays by a degree. But it does not accord with the Shannon principle and aliasing is introduced. A MAP （maximum a posterior）super- resolution restoration using optimal reciprocal cell is proposed to reduce aliasing. First, aliasing is modeled with dual grid in frequency domain. The optimal reciprocal cell with less noise and less alias is obtained by limiting the reciprocal cell with a weight function. The Huber-MRF （markov random field）prior is used in the super-resolution restoration with optimal reciprocal cell for edge-preserving. Experimental results indicate that the proposed method reduces the aliasing in recovery and obtains an effective image super-resolution restoration result.