L1范数的图像超分辨率重建改进算法

Improved Image Super-resolution Reconstruction Algorithm Based on L1-norm

介绍了超分辨率图像重建的数学模型和基于L1范数的超分辨率重建算法。针对在所观察到的低分辨率图像不足情况下的超分辨率重建,在L1范数重建算法框架下,提出了一种新的代价方程,在其中增加了关于丢失的低分辨率观察信息的保真度项和正则化项。该方法同时对高分辨率图像和丢失的观察信息进行迭代估计,并利用交替最小方法求解。实验结果表明,在获取低分辨率图像较少的情况下,提出的算法能够有效地改进重建的结果。

The mathematical model of super-resolution reconstruction and Ll-norm based reconstruction algorithm is introduced. And, here, the situation where some low-resolution images are missing is considered. A new cost function is presented under the Ll-norm reconstruction framework. The data fitting term and regularization term of the missed low-resolution images are added to the cost function. The alternating minimization method is used to estimate the high resolution image and missed low-resolution images. Experimental results demonstrate the effectiveness of the proposed method under condition of few low-resolution images observed.