ABSTRACT
MotionDSP makes software for super-resolution (and, more generally, video enhancement) that is used in various professional and consumer applications. This chapter discusses some of the design decisions made while developing and implementing super-resolution algorithms for the real world. It also details challenges, such as robustness, automation, speed (i.e., computational complexity), and specific limitations of a video imaging pipeline (acquisition speed, noise, and compression). Based on our experience, solutions to these challenges should be incorporated early into the design of super-resolution algorithms and products.
