循环-循环块相位掩膜矩阵压缩双透镜成像

压缩成像(CI)是压缩传感(CS)理论的一个重要应用领域。本文在循环矩阵和分块矩阵的基础上提出一种新的循环-循环块相位掩膜矩阵可压缩双透镜成像方法,实现对图像的随机获取。模拟实验结果表明,新的相位掩膜矩阵CI可以在显著减少测量的同时,有效捕获图像信息重建原始图像;对于64pixel×64pixel的标准Lena和Cameraman图像,在只有原像素数58.6%的无噪声测量情况下,获得的重建信噪比(SNR)分别超过26dB和30dB。新相位掩膜矩阵的研究为确定性测量在CI领域的应用提供了更多的支撑,在拥有原循环和循环确定性测量优点的同时还拥有自身的块结构特点,可以进一步减少物理实现成本,并为新的照相机设计提供若干理论、计算和技术支撑。

Compressive double-lens imaging using circulant-circulant-block phase mask matrices

Compressive imaging is an important application of the compressive sensing theory.This paper proposes a novel compressive double-lens imaging method with circulant-circulant-block phase mask matrices to implement image acquisition.Simulation results show that the novel phase mask matrices can effectively capture the information of image with a significant reduction of measurements.For the standard Lena and Cameraman images both with 64-by-64 pixels,the number of measurements is only 58.6% of all pixels,and the reconstructed signal-to-noise ratios exceed 26 dB and 30 dB,respectively.The new research provides more supports for deterministic measurement in the application of compressive imaging,which has more advantages than toeplitz and circulant matrices,reduces the difficulty and cost of the physical realization and provides some theoretical,computing and technical supports for new design of camera.