Quantum Assisted Image Registration

Quantum Computing has been invigorated by the introduction of the first commercially available computers. Quantum Annealers (QAs) like the D-Wave are now housing ∼5000 quantum-bits, and quantum computing, a potentially disruptive technology, is closer to becoming a reality than ever. Here we investigate quantum-assisted approaches to image registration of MODIS images using QAs. We present two approaches: (1) a mapping of image matching to a quadratic-unconstrained-binary optimization problem and (2) a machine learning approach that employs generative models trained using quantum annealing statistics. Results are shown and discussed.