A team of engineers at the University of Washington has developed a new generation of metasurfaces, resulting in a camera that is revolutionarily small. It is a square with a side of only 0.5 mm, like a grain of sand, but it produces an image of the same quality as a conventional video camera 500,000 times its size. Professional photographers do not agree, the novelty has a fair amount of flaws, but in this case it is secondary.
The main difference between this camera and a conventional one is that it has no lens system and no moving mechanisms at all. They are replaced by an array of 1.6 million “nanoposts” – structures of a specific shape that transform light according to a predetermined algorithm. The technology itself appeared a few years ago, but in the new version, the structure of the nanoposts has been radically transformed using machine learning and deconvolution methods based on neural functions.