基于第二代小波的序列图像超分辨率复原算法研究

介绍了第二代小波(SGWs)的基本原理和特点,在此基础上提出了低分辨率采样图像高分辨率(HR)重建算法.它不但具有结构简单,逆变换容易,可扩展性好的优点,并且克服了第一代小波(FGWs) 的不均匀采样和定义在有限区间上的难题,指出了小波变换时数据不规则性的处理方法,并利用硬/软阈值对小波系数降噪,使得重构后的图像具有最佳的PSNR和最低的重构误差.实验和仿真结果表明,与传统方法相比,该算法在提高了图像的分辨率同时PSNR最多能提高3 dB左右.

The Research for the Super-resolution Restoration Based on the Image Sequence with Second Generation Wavelets

On the basis of theories and characteristics of the second generation wavelet (SGWs), high-resolution (HR) reconstruction algorithm which is from a captured sequence of low-resolusion frames is proposed. It not only has a simple structure, a easy process of inverse transformation, the advantages of good scalability, but also overcomes bounded domains and irregular samples of first generation wavelet (FGWs). The approach of dealing with irregular data during the process of wavelet transform is introduced. By using hard /soft thresholding , the noise of wavelet coefficients is reduced , which makes the best PSNR and the lowest error of the reconstructed image. The result of experiment and computer simulations show that this method improves the resolution of images at the same time increase the PSNR up to about 3dB.