The advent of CCD/CMOS chip technology in digital cameras has brought improvements in their size, weight, cost, and reliability. However, the canonical configuration of such imaging chips performs poorly at capturing all the spatial detail available through a camera's optic. Additionally, they are noise-limited devices with upper bounds on photo-sensitivity.

Super-resolution is a technique of combining multiple displaced exposures of the same scene into a composite image. By treating each exposure as statistically independent, an aggregate estimation of the true scene can be generated with more samples and less variance than any single exposure, creating an image with dramatically improved spatial resolution that is consistent with physical and statistical law.

More interestingly, the assumption of statistical independence breaks down asymptotically for a variety of contemporary video signaling and data recording formats, and offers insight into the absolute performance limitation of super resolution techniques.