Super-Resolution

Super Resolution is a set of techniques that in some way enhance the Resolution of an imaging system, usually increasing image or video resolution.

The actual term was coined by scientists in 1984 as a mathematical breakthrough but did not become standard until about 1990.

All methods of super-resolution use the same basic idea, that of using information from the past and future 9 frames to create one upscaled image. Past methods used for Upscaling try to synthesize artificial details with questionable results. Super-resolution attempts to extract details from these past and future frames and then reconstruct other frames.

Super-Resolution is currently used for one of two tasks. Either upscaling whole video or as a way to extract a single Frame from video in the highest quality possible.