Seeing in a Different Light

January 2003
By Henry S. Kenyon

Quantum imaging opens new paths for optical sensors, holography, cryptography.

The U.S. military may one day obtain detailed reconnaissance imagery with laser light that has never touched a target. By using two laser beams and taking advantage of a unique characteristic of quantum mechanics that permits one beam to mirror the state of its twin, researchers are developing low-power systems that can measure, or illuminate, objects across a variety of frequencies, yet generate detailed pictures in the visible spectrum.

As the need for detailed military and commercial imagery grows, so does the need to reproduce that data accurately. Introducing devices that can flexibly illuminate objects in a variety of light frequencies would benefit both commanders and scientists, experts say.

Researchers at Boston University’s Quantum Imaging Laboratory, Massachusetts, are exploring potential applications for this new technology in areas such as imaging, holography and cryptography. 

Much of the research centers on a phenomenon called entanglement, which is a key principle of quantum mechanics. In the entangled state, two particles exhibit identical properties such as charge and frequency even though they are located in separate points in space. Scientists are using this characteristic in their optics experiments, explains Dr. Bahaa E. A. Saleh, Quantum Imaging Laboratory co-director and chairman of the university’s Electrical and Computer Engineering Department.

According to Saleh, quantum imaging begins with a source device that generates two laser beams. Each beam consists of a stream of single photons whose twin photons in the second beam are identical in frequency, direction and polarization. The photons are generated randomly, but because of the principle of entanglement they are identical when they are measured. If the direction and charge of one photon can be determined, it automatically indicates the corresponding state of its twin, he maintains.

This feature presents several advantages. If one beam illuminates an object, the other beam can generate its image. However, because of the properties of entanglement, the beam scanning the target does not have to be high-powered or high- resolution. The interrogating beam only has to return photons to a sensor. This target registration can be done with a simple one-pixel camera. High-resolution measurements are actually carried out by the second beam, which never encounters the target. “That is really the heart of the matter. It is the ability to interrogate an object with one beam and then measure the fact that the interrogation has taken place with very little dissolution. But then you take the other beam that is in the laboratory and has not traveled to the object and use a high-resolution camera to recover the information,” Saleh says.

This permits a kind of distributed image processing where the task of simple target illumination and signal collection is separated from the actual image processing itself, explains laboratory Co-Director Alexander V. Sergienko. “Paradoxically, the high-resolution processing is done with the beam that has never seen the object. The one that went to the object does not require sophisticated processing, it just needs to be detected,” he says.

This type of illumination could facilitate military target and data acquisition. One danger of ranging and illumination lasers is that many defensive sensor systems on military platforms can trace the beam back to the source. However, Sergienko notes, the interrogation beams used in quantum imaging are very weak and spread across several frequencies so they are not detected as a coherent point.  Sensors mistake these beams for background light.

However, he warns that there is a price to pay for being limited to low-level illumination. The low-power beam takes several seconds to illuminate and retrieve the data, making it unsuitable for certain time-sensitive applications, Sergienko explains. The quantum properties of entanglement generate the best images when low levels of photons are used. When a higher power beam is used, too many photons are generated, decreasing the resolution of the image.

Measurements on the first beam primarily indicate the direction and arrival time of the photons at the target object. This data, gathered from the twin beam, is the basis for generating an image of the object. “We are looking at the coincidence of the two photons. One photon tells us something has illuminated the object somewhere, but the other beam—the other photon—tells us where,” Saleh relates.

The two beams do not need to use the same wavelength.  For example, the interrogating beam can be in infrared, but the imaging beam may use the visible spectrum to generate a high-resolution image. This permits a relatively low frequency beam to illuminate an object and, as a result of entanglement, a picture can be recovered in high detail by the second beam.

Sergienko notes that this approach is particularly useful in the infrared spectrum because high-resolution infrared images are hard to obtain. By using sensitive infrared point detectors to record the arrival times of individual photons, an object can be measured by one beam in infrared, while the other beam generates a high-resolution image with an electronic camera. “In this way, your joined signal-to-noise ratio will be significantly higher than if you do a direct infrared measurement,” he explains.

The twin beams are generated with a nonlinear optical crystal, a device commonly used in laboratories to change a laser beam’s color. Once the beams are split and one is sent to interrogate an image, the only communication between the two occurs at the end of the process when the electrical signals from the two detectors are brought together to take measurements, Saleh explains. One advantage of using two beams is that the beam conducting the precise measurements can be located in a stable, secure place while the low-resolution beam interacts with the environment, he says. 

Quantum imaging also may be used to measure the amount of light an object reflects. Traditionally, exactly calibrated, known sources were required to obtain this data. However, if no known source is available, the laboratory’s two-beam method can provide this source because the photons for the beams are generated in pairs. “So if you keep one of these and measure it, then you know exactly what is hitting the object because you are measuring its twin. That is a way of calibrating. In other words, since you have two beams, you don’t need an additional calibration because the two beams must be exactly the same. It allows you to know what is hitting the unknown object,” Saleh says.

Since its founding in 1995, the Quantum Imaging Laboratory also has engaged in a number of other studies involving holography, polarization and cryptography. Quantum holography shares many of the same principles used in traditional holography.

Saleh notes that in traditional holography, a beam of light illuminates an object. Part of that beam passes through and is unaffected by the object while another part of the beam is scattered, creating an interference pattern. A hologram is the record of this interference pattern. Like traditional holography, the laboratory uses a reference beam that splits into two components when it scans an object. A single-pixel point detector instead of a high-resolution camera captures the image, and the second beam measures the object in detail. The second camera must be a high-resolution device for effective three-dimensional imaging, he says.

Researchers are investigating other applications for quantum imaging such as tomography, a technique used in medical systems to photograph tissue layers. Quantum tomography uses a beam to probe and image deep layers of tissue and organs in a body. Like other quantum systems, this is done with a ranging beam while the second beam performs the imaging.

The laboratory also is involved in quantum cryptography. Mirroring research at other laboratories (SIGNAL, December 2000, page 57), Boston University’s work involves exchanging key information in a manner that is unreadable by any eavesdroppers. Saleh notes that one difficulty of other quantum cryptography techniques is that the key information is sent down a fiber optic cable one photon at a time, which is done by attenuating a light source to only transmit single photons. However, a small possibility exists that two photons may be sent in a given pulse, and an eavesdropper could capture the second photon and recover the key information.

The laboratory’s cryptography approach generates photons in pairs. One photon is sent with the key data while the other resides at the source device. Because of entanglement, noting the time that the first photon was detected indicates that its twin exists, permitting data to be exchanged. This technique currently is being compared with other quantum cryptography efforts.

Working with Harvard University, the laboratory is developing an experimental quantum cryptography system. The project is funded by the Defense Advanced Research Projects Agency, Arlington, Virginia, and BBN Technologies, Cambridge, Massachusetts, the research arm of Verizon. According to Quantum Imaging Laboratory Co-Director Malvin C. Teich, the goal is to test quantum cryptography methods by transmitting entangled photons through fiber optic lines running underneath the streets of Boston and Cambridge to three stations at Boston University, Harvard and BBN. Teich expects the program to begin running experiments in the next two years.

Saleh notes that the results of the laboratory’s quantum research have been demonstrated to the government agencies funding the work, but some challenges must be overcome before these systems appear in operational equipment. A major hurdle is the inability to generate large numbers of entangled photons. This limitation increases the time required to capture an image. Low acquisition speed may be impractical for some uses, but applications with commercial value exist, he says. For example, given sufficient time, quantum tomography can capture deeper and higher resolution images. Therefore, this process may be used to scan samples of bone and tissue but is not suitable for scanning a live subject. 

The inability to create more entangled photons is an engineering challenge that is not insurmountable. “If someone invents a methodology for generating these entangled photons in large numbers, then I would say that in a very short period of time, many of these applications would see commercial use,” Saleh says.

 Additional information on Boston University’s Quantum Imaging Laboratory is available on the World Wide Web at