对不同分辨率图象进行空间分辨率改善的频域方法

图象空间分辨率的改善在遥感图象和计算机视觉的一些应用中非常必要。已有的一些空间分辨率改善的频域方法是针对多幅相同空间分辨率图象的频谱解混叠方法。本文从分辨率低的欠采样图象会导致相应空间频率域频谱混叠的理论出发，给出了多次欠采样图象在频率域混叠的更一般的公式，并给出一种针对不同分辨率图象解频谱混叠的逐行迭代方法。最后通过计算机仿真实验对本文所述理论进行了验证。实验表明，该迭代解法在有噪声情形下也具有很好的收敛性。

A Frequency Method for Image Resolution Improvement from Multiframes of Different Resolutions

In the applications such as remote sensing and computer vision, to reconstruct a high-resolution image from multiframes of low-resolution images is required. Former work was based o n the same r eso lution fr ames. T his paper is based on the theorem of spectral aliasing resulting from undersampled low-resolution image, and a general formula of the spectral aliasingis proposed for the problem of different resolutions. A row-action method even for noisy and blurred frames are given in the DFT domain as well. Computer simulations demonstrate the performance of the algorithm.