"Ghost imaging", also called two-photon imaging or correlated imaging, is an imaging technique that reconstructs spatial information of an object by using joint detection of paired photons.
Overview
Conventional optical observation measures the intensity distribution of the optical field. Correlation-based optics measures intensity correlations of the optical field, while most imaging techniques use first-order field information (intensity and phase). Classical ghost imaging exploits second-order correlations of the field, which represent statistical correlations of intensity fluctuations. Ghost imaging has been applied in radar, remote sensing, photography, X-ray imaging, neutron imaging, electron imaging, cold-atom imaging, fluorescence microscopy, acoustic sensing, and 3D printing.
History
The spatial nonlocality of entangled photon pairs, a conclusion related to the Einstein-Podolsky-Rosen (EPR) paradox, attracted wide attention and led to research in quantum information. Early roots of ghost imaging trace back to experiments in the 1950s and 1960s: the Hanbury Brown and Twiss (HBT) intensity interferometry experiments were related to these ideas. Those experiments addressed atmospheric disturbance problems in traditional stellar interferometry and revealed coherence properties in quantum optics.
Later, Russian researchers used similar methods to display diffraction fringes of an object's edges on an optical path that did not contain the object. In 1993, Brazilian researchers experimentally showed that using entangled thermal-like sources and coincidence counting could recover Young's interference fringes that would otherwise disappear due to decoherence.
In the mid-1990s, with successful preparation of entangled photon sources, ghost imaging reappeared as "quantum imaging." One photon of an entangled pair interacts with an object and is detected by a bucket detector without spatial resolution, while its partner photon is detected by a spatially resolving detector. Correlating the two detectors' results can produce an image of the object.
The quantum interpretation led some researchers to view entanglement as essential, while others argued it was not strictly necessary. The debate continued until 2002, when Bennink and colleagues in the US demonstrated ghost imaging using a classical, non-entangled source. Subsequent theoretical analyses based on classical statistical optics and coherence theory showed that thermal light can produce ghost imaging through classical correlations. From that point, ghost imaging was no longer limited to quantum imaging and began to find broader applications.
Research into nonlocal quantum imaging then expanded rapidly. The term "ghost imaging" reflects the surprising fact that neither detector alone forms a direct image of the object, yet correlating the two detector outputs reconstructs the object's image, as if two unaware artists with closed eyes cooperated to draw a precise portrait.
Principle
First, consider classical imaging. Light from an object passes through an optical system to form an image (for example, the eye, a camera, or a lens). Typical components are the light source, the object, and the optical system.
Ghost imaging works as follows. Light from a source passes through a random mask (for example, a rotating ground glass) and is split by a beamsplitter into an object arm and a reference arm. In the object arm, the speckle field illuminates the object; the transmitted or reflected signal is recorded by a bucket detector that only measures total intensity without spatial resolution. Simultaneously, the speckle field that did not interact with the object is recorded by a CCD camera in the reference arm. One measurement consists of the bucket detector readout and the reference camera frame; after N samples, the object image can be reconstructed from correlations. In practice, a detector D1 placed outdoors and a detector D2 sampled indoors could, after correlation, reconstruct the outdoor scene without directly observing it.
Advantages
Ghost imaging's indirect approach yields several advantages over conventional imaging. Because the bucket detector collects all transmitted or backscattered light without spatial resolution, ghost imaging is more robust to atmospheric disturbances such as clouds, smoke, and haze, enabling clearer images under adverse conditions. Collecting all object light into a single detector also avoids distributing photon energy across pixels of an array detector, improving signal-to-noise ratio; consequently, ghost imaging can operate under extremely low illumination. This approach can also be applied to radiation-hazardous imaging, such as X-ray imaging, enabling low-dose X-ray ghost imaging.
Experimental results indicate that under weak illumination, ghost imaging can achieve higher signal-to-noise ratio than traditional transmission imaging. For the same SNR, ghost imaging can significantly reduce radiation dose during imaging. This experimental validation of ultra-low-dose ghost imaging supports further development toward 3D X-ray ghost imaging and biomedical applications.
Because ghost imaging can penetrate scattering media, it has also produced notable results in remote sensing and surveillance.
Developments
1. X-ray Ghost Imaging
To address high coherence requirements and the difficulty of making X-ray lenses, ghost imaging uses incoherent sources to achieve lensless diffraction imaging. This enables compact desktop X-ray diffraction imagers and advances X-ray applications in nanotechnology, life sciences, and remote sensing.
2. Ghost Radar
From an imaging perspective, radar constructs images from received echoes. Ghost radar obtains spatial and three-dimensional information by correlating target echo signals with transmitted signals. Unlike traditional imaging systems, ghost radar uses statistical image properties to reduce sampling requirements and increase imaging speed while offering super-resolution. For example, a conventional imaging grid of 100×100 points requires 10,000 measurements, though effective information may only occupy a smaller region. Ghost radar avoids many redundant measurements and can reconstruct complete information from far fewer samples. As an optical remote sensing technique, ghost radar combines long-range detection capabilities with high-resolution imaging similar to flash imaging radar.
For instance, a ghost imaging device mounted under an aircraft can image ground targets and produce results where different colors represent height information; converting that data yields three-dimensional target reconstructions.
3. Ghost Imaging Camera
Ghost imaging cameras can record spatial information like conventional cameras and may also capture additional optical dimensions such as spectrum and polarization in a single exposure. Ghost imaging has been applied successfully in fluorescence microscopy, relying on ghost imaging camera techniques.
Live-cell imaging is critical for understanding biological mechanisms and dynamics. Optical microscopes are primary tools, but spatial resolution is limited by optics to around 200–300 nm. Super-resolution methods such as stimulated emission depletion microscopy (STED), structured illumination microscopy (SIM), photoactivated localization microscopy (PALM), and stochastic optical reconstruction microscopy (STORM) overcame diffraction limits but face trade-offs between temporal and spatial resolution and risk of photodamage from high-intensity illumination. As a result, current super-resolution fluorescence techniques still struggle to observe fast nanoscale dynamics inside living cells in real time.
