基于压缩感知的图像压缩抗干扰重构算法

An anti-interference reconstruction algorithm of image compression based on compressed sensing

针对传统图像变换压缩方法压缩的图像经无线信道传输时受高斯随机干扰导致重要变换系数失真出现重构图像局部内容缺失的现象,本文根据压缩感知(CS)信号分量具有同等重要性的特性,理论分析了去除失真CS信号分量以抵御干扰的可行性,提出一种基于CS的图像压缩抗干扰重构算法。算法首先假定已知受高斯随机干扰的比特所对应的CS信号分量的位置,然后根据这些位置确定新的CS信号和重构矩阵,再进行阈值迭代重构。仿真结果表明,本文算法在低误码率(BER)下得到精确重构的图像,在高BER下得到图像内容无缺失仅全局质量小幅下降的重构图像。因此,基于CS的图像压缩抗干扰重构算法能够较好地克服变换压缩方法以及阈值迭代重构算法抗干扰能力低的不足,从而为图像无线传输抗高斯随机干扰问题提供一种可行的解决方案。

When the images compressed by traditional transformation compression algorithms are trans- mitted over wireless channels,if Gaussian random interference causes the loss of the crucial transforma- tion coefficients,the contents of the reconstructed images will be lost obviously and this will reduce the accuracy of the subsequent detection and recognition results greatly. In order to solve this problem, based on the characteristics of equal importance about each compressed sensing component, this paper first an- alyzes the feasibility of resisting the interference by removing the distorted compressed sensing signal components, then proposes an anti-interference image reconstruction algorithm. This algorithm first con- firms the new compressed sensing signals and the new reconstruction matrix based on the locations of the corresponding compressed sensing signal components of the Gaussian-interfered bits, and then recon- structs the original images employing the iterative threshold algorithm. The simulation results demon- strate that our algorithm can reconstruct exact images at low bit error rates, and inexact images whose qualities are slightly lowered without loss of local contents at high bit error rates. As a result, our algo- rithm can overcome the deficiency in anti-interference ability of transformation compression algorithms and the iterative threshold algorithm, thus proposes a feasible scheme for the anti-interference problem that arises in wireless image transmission.