Laser Technique Allows Noiseless Image Amplification

Dr. James P. Smith

Investigators at Northwestern University have demonstrated that it is possible to enhance images of faint objects without introducing graininess. Prem Kumar, a professor in the department of electrical and computer engineering, and colleagues at the university used a phase-sensitive optical parametric amplifier to increase very faint coherent signals without noise, and applied this "noiseless" amplification of time-varying signals to images -- signals that vary in both space and time. Their achievement was reported in the Sept. 6 issue of Physical Review Letters.

The amplifier is based on a potassium titanyl phosphate (KTP) crystal that converts green photons from a pump laser into pairs of infrared photons. The device can be made to amplify an IR signal because it emits more IR photons than would be produced by the pump beam alone when the signal and green pump beams simultaneously hit the crystal.

The experimental setup -- a KTP crystal pumped with a frequency-doubled beam (532 nm) from an Nd:YAG laser -- used a double-slit pattern as the image to be amplified. A 1064-nm signal beam from the laser passed through the slit pattern, combined with the 532-nm pump beam and passed into the KTP crystal. An InGaAs photodetector then scanned the resulting image. A piezoelectric transducer in a feedback loop continually adjusted the pump beam's relative phase to the signal beam to obtain maximum amplification.

If the IR signal is uncorrelated with the pump beam, IR photon pairs interfere with the stimulated pairs and create a noisy signal. When the phases of the two beams were synchronized, however, the stimulated pairs were in phase with the signal, and noiseless amplification occurred.

Kumar believes that this method has many potential applications. "The technique may be applicable to many situations where weak coherent signals must be preamplified," he said, "such as in laser radar and in imaging through highly scattering systems such as laser medical imaging through tissue. The important thing, however, is that the signal must be coherent light. The technique will not work for starlight or other incoherent sources."

Kumar said the ideas behind the technique are not new. "We have no plans to apply for a patent. There have been a lot of inquiries from people with ideas for applying the technology, and we hope to work on some of these," he said.