基于MAP算法的图像超分辨率重建

MAP-Based Image Super-resolution Reconstruction

近年来图像的超分辨率重建已经成为人们广泛研究的热点。本文提出了一种从多幅低分辨率欠采样图像中重建出一幅高分辨率图像的重建方法。该方法基于MAP框架,用迭代方法得到最优化解。在每次的迭代过程中利用上次迭代得到的重建图像的有用信息来不断调整迭代参数,不断的循环迭代,最后求解出重建图像的最优解。实验结果证明,该方法有效,它不仅能在迭代过程中自动选择和更新调整参数,并且能得到期望的高分辨率重建图像。

Super-resolution image reconstruction has been one of the most active research fields in recent years. In this paper, a new super-resolution algorithm is proposed to the problem of obtaining a high-resolution image from several low-resolution images that have been sub-sampled. The algorithm is based on the MAP framework, solving the optimization by proposed iteration steps. At each iteration step, the regularization parameter is updated using the partially reconstructed image solved at the last step. The proposed algorithm is tested on Lena images. The results of the experiments indicate that the proposed algorithm has considerable effectiveness in that it can not only make an automatic choice and renew the regularization parameter, but also can get the high resolution reconstruction image expeetedly.