基于提升小波双正则化超分辨率算法

Bi-regularized Super-resolution Algorithm Based on Lifting Wavelet

基于已有算法的不足,提出了双正则化约束算法。该算法引入图像平滑性约束和点扩散函数平滑性约束,进一步缩小了重建的解空间。在提升小波域上实现算法求解,减少了算法的运算量,加快了迭代收敛速度,降低了对图像平稳性的要求。实验结果表明,该算法能有效地重建超分辨率图像。

In this paper,an effective algorithm for super-resolution image reconstruction is proposed.It uses bi-regularized function to reduce the solution space.The algorithm is carried on the lifting wavelet domain to speed up convergence rate.The simulation results show that this algorithm is effective.