Local self-similarity based image super-resolution reconstruction algorithm

doi:10.11834/jig.20110526

Views : 0 下载量: 351 CSCD: 0

PDF
Export
Share
Collection
Album

Local self-similarity based image super-resolution reconstruction algorithm
Vol. 16, Issue 5, Pages: 720-725(2011)
Online First：23 May 2011，

Published：2011
DOI： 10.11834/jig.20110526
稿件说明：

移动端阅览

DOI：

Local self-similarity based image super-resolution reconstruction algorithm[J]. Journal of Image and Graphics, 2011, 16(5): 720-725. DOI： 10.11834/jig.20110526.

摘要

图像超分辨率重构是指由低分辨率图像来获得高分辨率图像的过程。为了能够有效地重构出高分辨率图像，提出一种基于图像局部自相似性的超分辨率快速重构算法。该算法首先利用四叉树分割的知识对低分辨率图像进行自适应分块；然后利用低分辨率图像和高分辨率图像在局部区域内的自相似性，由最小二乘方法在各个局部区域自适应的选择插值所需的参数，从而在各个局部区域内进行插值；最后运用小波域的投影算子对插值得到的高分辨率图像进行全局优化，得到最终的高分辨率图像。实验结果表明，由该算法重构的高分辨图像有很好的视觉效果和峰值信噪比。

Abstract

Image super-resolution refers to reconstruction of a high resolution image from one or a set of blurred low resolution images. This paper only pays attention to the kind of reconstruction from one blurred low resolution image. Many methods have been developed for this kind of reconstruction

most of which are MAP methods and interpolation methods. This paper proposed a new interpolation method. The proposed method used the quad tree segmentation to partition the low resolution image

the edge-directed interpolation to each segmented band of the low resolution image

and a wavelet projection to optimize the high resolution image obrained from the local interpolation. The experiment used the peak signal to noise ratio (PSNR) to compare the reconstructed image with the original image. And the results showed that the PSNR and visual effect of the high resolution image reconstructed with the proposed method were very good.