Through eyes or camera, the world is observed by people with light intensity measurements.

Microscopically, this means photons with certain frequencies and propagation modes are counted electronically. Charge-transfers are induced during the process to record the intensity profile of the optical modes. The phase information are lost.

Do we have a phase-sensitive detector to resolve the optical phase? It turns out that any phase measurement must be relative. Therefore, a phase-sensitive detector requires a "local oscillator" to provide a reference phase. In wireless communication, we do this all the time. It is technically feasible to have a stable reference oscillator in your mobile phone, either 2G or 5G or 6G. Can we do the same thing in optical communication? In principle yes, but rarely practical so far for a list of reasons. A major problem is associated with the short optical wavelength of light which make the phase of optical modes extremely sensitive to perturbations (Doppler effect, for example). The sensitivity means optical local field is difficult to distribute by independent sources (Although such an achievement would be extremely useful, such as for intersteller optical communication). More often people do a "beamsplitter" trick: To split light into a "reference path" with complex amplitude Er and a "imaging path" to probe certain samples. The mode profile Er is simple enough and assumed as known. Within a ``coherence length'', the Es transmitted through the imaging path can interfere with Er on a camera to record a conj(Er)Es+c.c. term in intensity, from which people try to reconstruct Es. The idea is known as holography.

Holographic microscopy was proposed by Dennis Gabor. It is a general principle of coherent imaging that records not only intensity, but also phase of waves. Here the waves can be for photons or light, as well as for electrons and neutrons. Nowadays, the ``new'' microscopic principle by Gabor is widely used in material science, surface science, and biological science. 

But the method is rarely exploited in atomic physics for imaging purpose. 

Our work so far demonstrates that holographic microscopy, when constructed properly, can be a powerful tool for retrieving spatial-dependent spectroscopic signal of laser-cooled atoms. The signals are ``complex'', containing both the absorption and phase shift to light by the atoms. We are able to retrieve the ``phase angles'' of the complex signals, and associate them with collective response of confined atoms. Our technique may impact ultra-cold research associated with precision measurements, quantum sensing, and quantum gas microscopy. The technique may also contribute to the development of coherent imaging and sensing with light.

We are also working on improving the imaging speed (toward e.g. 1-frame/sec level) and sensitivity (toward single-atom level). 

The project involves aspects of atomic physics and quantum optics, digital holography, as well as novel ideas in coherent imaging such as by Gaussian beam decompositions.

We are looking for new team members. If your major is physics, interested in computer automation, numerical simulation and data analysis, you might like our project -- please drop by for a discussion. 

面向相干原子光谱成像的精密全息显微技术

无论是用双眼还是照相机，人类总是通过测量光强来观察世界。

微观地来说，这意味着具有一定频率和传播模式的光子被电子计数。在这个过程中，电荷转移被用来记录光场的强度分布，而相位信息则会丢失。

我们有可以测量光相位的相位敏感探测器吗？事实上，由于任何相位测量都是相对的，相位敏感探测器需要一个 本地振子 来提供参考标准。在无线通信中，我们一直是这么做的。在你的移动电话里有一个稳定的参考振荡器，无论是2G还是5G或6G，这在技术上是可行的。我们能在光通信中做同样的事情吗？原则上也可以，但由于一系列原因，该想法并不实用。其中一个主要问题与光的波长有关，由于光的波长短，其相位对外界扰动极其敏感（例如以多普勒效应）。此类敏感性意味着光场局部振子很难由独立的光源提供（尽管如果能这么做会非常有用，如星际光通讯）。更多的时候，人们用一个“分光镜”技巧：把光分成一个具有复振幅Er的“参考光”和另一个用于探测样品的“成像光”。Er的模式分布很简单，可假定已知。在“相干长度”内，通过成像路径传输的Es可以与照相机上的Er发生干涉，因此强度分布可提供干涉项conj(Er)Es+c.c.，从中人们可尝试重构出待测场Es。这个技术被称为光学全息。

全息显微最早是由Dennis Gabor提出的，是一个普适的相干成像原理来同时记录波场的强度和相位。这里的波可以是光子，也可以是电子和中子。如今，Gabor的“新”显微原理被广泛用于材料科学、表面科学和生物科学。

但该方法很少被用于原子物理学中来成像。

迄今为止，我们的工作表明，全息显微镜，如果构建得当，可以成为探测被激光冷却原子空间分辨光谱信号的有力工具。且这些信号是“复数”，同时包含原子对光的吸收和相移。我们能够检测出该复信号的“相角”，并将其与囚禁原子的光协同相互作用联系起来。我们的技术可能会对精密测量、量子传感和量子气体显微镜有关的超冷研究产生重要影响。我们的新技术也有希望促进相干成像和传感技术的发展。

我们还在努力提高数字全息成像速度（致1帧/秒水平）和灵敏度（致单原子水平）。

该项目涉及原子物理学和量子光学、数字全息，也涉及相干成像的全新想法，包括通过高斯光束分解复杂光场的路线。

我们正在寻找新的团队成员。如果你的专业是物理学，对计算机自动化、数值模拟和数据分析感兴趣，你可能会喜欢我们的项目——请来信来函讨论各种可能。