压缩成像是压缩感知理论最重要的研究领域之一.在分析压缩成像中实际测量矩阵与测量值约束的基础上,提出一种基于4-f光学架构的物理可实现的频域相位编码压缩成像方法.该方法利用两路光学架构之间的补偿实现相位编码压缩成像中测量值的非负记录,然后从该测量值精确恢复原图像,解决压缩成像应用中理论与实际物理约束之间不一致的问题.该方法可以单次曝光获得充分的测量值精确重建原图像,不需要其它附加信息,是压缩成像物理实现的一种非常可行的方案.模拟实验证明该方法可以有效地实现图像的压缩测量与超分辨率重建.
Compressed Imaging(CI) is one of the most important research area of compressed sensing.Analyzing the constraints on actual measurement matrix and measurement values in CI,a frequency domain phase encoding CI method is proposed,which can be physically realized based on 4-f optical architecture.This method exploits two-way optical architecture compensated for the phase-encoding CI to implement the value non-negative for recording,and then accurately recover the original image from the measured values,to resolve inconsistencies between the theoretical requirements and physical constraints in CI.With measured values can be obtained in a single exposure,such method can reconstruct the original image precisely without additional information,and is a very practical scheme for physical realization of CI.Simulation experiments demonstrate that our proposed method can effectively capture compressed measurements of image and achieve super-resolution reconstruction.