Wavelet sparse transform optimization in image reconstruction based on compressed sensing

3rd International Conference on Advances in Energy, Environment and Chemical Engineering

能源学;生态环境科学;化学工业

Ziran, Wei^1,2 ; Huachuang, Wang^1 ; Jianlin, Zhang^1

Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu

610209, China^1

University of Chinese Academy of Sciences, Beijing

100039, China^2

关键词: Optimization method;    Original signal;    Peak Signal to Noise Ratio (PSNR);    Reconstructed image;    Reconstruction accuracy;    Reconstruction image;    Sparse transform;    Wavelet coefficients;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/69/1/012165/pdf
DOI  :  10.1088/1755-1315/69/1/012165

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

The high image sparsity is very important to improve the accuracy of compressed sensing reconstruction image, and the wavelet transform can make the image sparse obviously. This paper is the optimization method based on wavelet sparse transform in image reconstruction based on compressed sensing, and we have designed a restraining matrix to optimize the wavelet sparse transform. Firstly, the wavelet coefficients are obtained by wavelet transform of the original signal data, and the wavelet coefficients have a tendency of decreasing gradually. The restraining matrix is used to restrain the small coefficients and is a part of image sparse transform, so as to make the wavelet coefficients more sparse. When the sampling rate is between 0. 15 and 0. 45, the simulation results show that the quality promotion of the reconstructed image is the best, and the peak signal to noise ratio (PSNR) is increased by about 0.5dB to 1dB. At the same time, it is more obvious to improve the reconstruction accuracy of the fingerprint texture image, which to some extent makes up for the shortcomings that reconstruction of texture image by compressed sensing based on the wavelet transform has the low accuracy.