使用提升小波进行超分辨率图像重建

Superresolution image reconstruction with lifting scheme wavelet transform

介绍了提升小波变换,研究了提升小波的优点,提出了使用提升小波进行超分辨率图像重建的方法。对原图进行旋转、平移、加噪并进行下采样,得到低分辨率图像序列。选取阈值,对低分辨率图像序列进行提升小波前向变换,然后通过提升小波逆变换得到重建以后的超分辨率图像。与传统的几种重建方法进行比较,实验结果表明,本文的方法能够使重建图像有更好的信噪比,而且在像素较大和低分辨率图像数目较多的情况下,有更快的重建速度。

This paper introduces the lifting scheme wavelet transform and the way using it in the superresolution（SR） image reconstruction field.A series of low resolution （LR） images can be acquired from rotation,shifting,adding noise and down sampling of one image.Using forward lifting wavelet transform to LR image sequence with chosen threshold,then the reconstructed SR image can be acquired from the inverse lifting wavelet transform.Compared with the traditional reconstruction methods,the experiment result shows that the reconstructed imaged using this method has a higher SNR and if the image pixel or the number of the LR image sequence is very large this method has a higher speed.