一种自适应超分辨率图像重建方法研究

Research on an Adaptive Super-resolution Image Reconstruction Method

针对图像去噪的问题,提出了一种自适应范数及正则化参数的图像重建方法。首先,考虑到退化图像不仅含有高斯噪声,而且含有拉普拉斯噪声,利用最大似然估计的思想估计高斯噪声和拉普拉斯噪声的标准差;其次,由于在图像重建过程中,噪声分布会发生变化,为此,构造基于统计量的高斯和拉普拉斯权重函数,整合L_1、L_2范数,设计一种自适应加权函数;最后,结合自适应正则化参数方法,设计了一种自适应L_1、L_2范数及正则化参数的图像重建方法。实验结果表明,提出的方法对含有混合噪声的不同图像具有比较理想的重建效果。

Aiming at the denoising problem in images,an image reconstruction method based on an adaptive norm and regularization parameter is proposed.Firstly,as the degraded images have Gaussian noise and Laplacian noise,the idea of the maximum likelihood estimation is adopted to estimate the standard deviation of Gaussian noise and Laplacian noise;Secondly,in reconstruction process,the distribution ratio of noise could change,for this purpose,the weighting function of Gussian and Laplace is constructed by statistics,which integrates L1 and L2 norm,and designins an adaptive weighting function;Finally,with adaptive regularization parameter method,designs an image reconstruction method based on an adaptive L1-L2 norm and regularization parameter.The experimental results confirm the superiority of the proposed method for different images with mixed noise and outliers