Optical imaging is ubiquitous; it is an essential part of vision and influences how we record and display information. In classical imaging the photons illuminating the scene must interact with the object, travel from the object to the image plane, transferring information by virtue of the “position correlations” established by the optical system itself. For example, a lens may map the object plane to the image plane, point to point, thus correlating the positions. When this is done properly we retrieve a “sharp” image, while when the positions are not mapped precisely, i.e., poor position correlations, we find a “blurry” image. This type of imaging with man-made devices has been around for many many centuries.
By exploiting the quantum world it is possible to establish the necessary position correlations by virtue of quantum entanglement. For example, two photons may be entangled with one another, sharing quantum correlations at the source. One photon is then sent to the object and the other, which has never seen the object, to the imaging detector. By measuring the two in coincidence, thus extracting their quantum correlations, we can reconstruct a quantum image of the object. At Wits we are exploring the use of quantum technologies to image with very few photons, probing light-sensitive structures. By adding in machine learning and artificial intelligence, we can make imaging intelligent for fast recognition of objects.