滤波对计算鬼成像重构质量的研究

Research on the Quality of Computational Ghost Imaging by Filtering

通过对传统鬼成像技术中的双光路系统进行极大简化,计算鬼成像技术只使用一个不具备空间分辨能力的桶探测器即可实现目标图像重建。此方法不仅降低了成像成本,还可以在特殊波段成像。然而,采样数量大以及成像质量差仍然是很大的问题。针对成像质量差问题,通过理论分析、数值仿真研究了滤波对计算鬼成像重构质量的影响,并进一步比较了不同散斑下重构图像的信噪比,从数据上滤波对重构图像的质量看还是有一定提升效果的,并且在加入噪声后这种效果依然存在,还找出了哈达玛散斑按能量排序的规律,排序后的哈达玛散斑重构效果要好于其他散斑。

By greatly simplifying the dual-rail system in traditional ghost imaging technology,Computational ghost imaging only requires a single bucket detector without spatial resolution capabilities to reconstruct the target image.This approach not only reduces imaging costs but also enables imaging in specific spectral bands.However,the issues of large sampling quantity and poor imaging quality remain significant challenges.To address the problem of poor imaging quality,this paper investigates the impact of filtering on the reconstruction quality of computational ghost imaging through theoretical analysis and numerical simulations.Furthermore,the signal-to-noise ratios of reconstructed images under different scattering conditions are compared,which greatly contributes to the practical application of computational ghost imaging.