A new algorithm merges quantum and classical information to produce high-quality images

Researchers from Colorado State University and the Colorado School of Mines have devised a new computational imaging strategy that harnesses the best of both the quantum world and the classical world. They have developed an efficient and powerful algorithm that combines quantum information and classical information to produce high-quality images. The results of their study were published on December 21, Intelligent computing.

Recently, the quantum properties of light have been exploited to enable super-resolution microscopy. Although quantum information offers new possibilities, it also has its own limitations.

The researchers’ approach is based on classical and quantum correlation functions obtained from the number of photons, collected from spatially structured illumination-illuminated quantum emitters. The number of photons is processed and converted into ascending ordered signals, containing ascending spatial frequency information. However, information with higher spatial resolution suffers from a reduced signal-to-noise ratio at increasingly larger orders of correlation.

To solve the problem, the researchers developed a algorithm what they call “super deconvolutional images.” The goal of this algorithm is to strongly combine classical information, contributing high signal-to-noise ratio but low spatial frequency information, with quantum informationThis contributes to the reduction of the incoming signalnoise ratio but high spatial frequency information.

To that end, the researchers combined anti-bundle correlation images obtained from single quantum emitters with classical images obtained from the same photon counting data. Anti-bundling is a unique feature of single quantum emitters that can be used for super-resolution imaging.

The strategy benefits from overlap in the lower spatial frequency regions. The algorithm exploits the self-consistency requirement of overlapping measured spatial frequency information to trigger lower SNR (signal over noise) information at spatial frequencies, the researchers explain. high due to the contribution of the quantum image”.

Super decoding image algorithm leads to a significant increase in space Regularity content, faster speeds and higher resolutions, along with much better mean squared errors in the reconstructed images. The researchers believe their algorithm will be widely used because it is suitable for many different imaging contexts.

Powered by Intelligent Computing