Analysis of active superresolution imaging techniques for realtime applications

Abstract

In the field of superresolution (SR) imaging, most attention has been paid to reconstruction algorithm, but this kind of passive SR imaging technique with heavy computational cost and registration error cannot satisfy some realtime applications. Here we emphasize on active SR imaging technique and the warping and blurring between successive low resolution image frames are accurately controlled to a simple form during image acquisition. We present four kinds of active SR imaging techniques including two novel and promising ones. Using a spatial discrete observation model, we will unify all their reconstruction algorithms into a spatial domain based least square estimation algorithm, and we will also compare these four methods from theoretical and practical points of view. The simulation experimental results demonstrate their validities in aliasing reduction and resolution enhancement.