基于降质信息估计的盲图像复原算法

Blind Image Restoration Based on Degradation Information Estimation

针对低质量图像的复原重建问题,提出了一种基于降质信息估计的盲图像复原算法。该算法主要包括噪声估计网络、模糊核估计网络和重建网络3部分。首先分别通过噪声估计网络和模糊核估计网络,对图像噪声水平和模糊核进行估计;其次,将估计所得噪声水平和模糊核作为降质信息,并联合待处理的低质量图像一起输入重建网络,以帮助获得更好的重建效果。实验结果表明,该算法能获得良好的重建结果,无论是在客观评价指标上还是主观视觉效果上都有较好的性能。

Aiming at the problem of low quality image restoration, a blind image restoration algorithm based on degradation information estimation is proposed. The algorithm mainly includes three parts: noise estimation network, fuzzy kernel estimation network and reconstruction network. Firstly, the image noise level and the fuzzy kernel are estimated by the noise estimation network and the fuzzy kernel estimation network respectively. Then, the estimated noise level and the fuzzy kernel are used as degradation information, and fed into the reconstruction network together with the low-quality image to obtain better reconstruction effect.Experimental results indicate that the proposed algorithm can achieve fairly good restoration performance.Compared with state-of-the-art methods, it gets better results in both objective indexes and visual effects.