一种自适应压缩感知图像重构算法

An Image Reconstruction Algorithm Based on Adaptive Compression Sensing

为解决稀疏度未知图像的重构问题,基于小波变换提出了一种自适应压缩感知算法。详细论述了压缩感知算法的基本原理。利用小波变换对图像进行分解,得到高频子带和低频子带。考虑到各部分的稀疏性,仅对高频子带进行测量,保持低频子带不变,并将图像分为高高、高低、低高、低低等模块进行处理。针对稀疏度未知的情况,采用稀疏度自适应匹配追踪算法分别对包含在高频子带各部分中的高频系数进行恢复,通过小波逆变换进行图像重构。最后,以Pepper和Baboo图像为例进行了仿真实验。仿真结果表明,所述算法可较好地实现图像重构,能够提高图像重构质量,验证了所述压缩感知算法的有效性。

In order to solve the problem of image reconstruction with unknown sparse degree, an adaptive compression sensing algorithm is proposed based on wavelet transform. The basic principle of compression sensing algorithm is described in detail. The image decomposition is carried on by using wavelet transform. The high-frequency and low-frequency sub-bands are obtained. Considering the sparse property of each part, the high frequency sub-band is measured while the low frequency sub-band is unchanged. And the image is divided into high-high, high-low, low-high and low high blocks for processing. Aiming at the condition with unknown sparse degree, the high-frequency coefficients contained in all parts of the high-frequency sub-band are recovered based on the sparse degree adaptive matching pursuit algorithm. The image reconstruction is realized by using wavelet inverse transform. Finally, take Pepper and Baboo images as examples, the simulation experiments are carried out. The simulation results show that this algorithm can well realize the image reconstruction and improve the quality of image reconstruction. The effectiveness of the compressed sensin- algorithm described is also verified.