Optical time-stretch (OTS) imaging has shown significant advantages in many applications, such as high-throughput cell screening, for its high frame rate and the capability of continuous image acquisition. Unfortunately, its application in practice is fundamentally limited by the pressure on data transmission and storage caused by the extreme throughput. Compressed sensing (CS) has been considered a promising approach to solve this problem by recovering the signal at a sampling rate significantly lower than the Nyquist sampling rate. However, the temporal stretch and compression processes, which are currently realized with two sections of dispersive media with complementary GVD, make the system complex and introduce notable power loss to the system, leading to a low signal-to-noise ratio (SNR). In this work, we propose and demonstrate all-optical Fourier-domain-compressed time-stretch imaging with low-pass filtering. Only one section of dispersive media is needed to temporal stretch the pulse, while the compression of the stretched pulses is achieved with low-pass-filtering after optical-electrical conversion. Therefore, the structure of the system is notably simplified, and the SNR or the quality of the images can be improved. The principle of this method is theoretically analyzed, and its performance is experimentally demonstrated. The results show that our system can image flowing cells at a high flowing speed of 1 m/s, with a spatial resolution of 2.19 μm on condition that the original data are compressed by 80%. Our method can boost the application of OTS imaging in the fields where high-throughput and real-time imaging is required.

中文翻译：

---

具有低通滤波的全光学傅里叶域压缩时间拉伸成像

光学时间拉伸（OTS）成像因其高帧速率和连续图像采集的能力而在许多应用中显示出显着的优势，例如高通量细胞筛选。不幸的是，其在实践中的应用从根本上受到极限吞吐量带来的数据传输和存储压力的限制。压缩感知（CS）被认为是解决该问题的一种有前景的方法，它可以以明显低于奈奎斯特采样率的采样率恢复信号。然而，目前通过具有互补GVD的两段色散介质实现的时间拉伸和压缩过程使得系统变得复杂，并且给系统带来显着的功率损耗，从而导致低信噪比（SNR）。在这项工作中，我们提出并演示了具有低通滤波的全光学傅立叶域压缩时间拉伸成像。只需要一段色散介质即可对脉冲进行时间展宽，而展宽脉冲的压缩则通过光电转换后的低通滤波来实现。因此，系统的结构显着简化，并且可以提高信噪比或图像质量。从理论上分析了该方法的原理，并通过实验证明了其性能。结果表明，在原始数据压缩80%的情况下，我们的系统可以以1 m/s的高流速对流动细胞进行成像，空间分辨率为2.19 μm。我们的方法可以促进OTS成像在需要高通量和实时成像的领域中的应用。