A sur-pixel scan method for super-resolution reconstruction

Abstract

The key technique in super-resolution reconstruction is sub-pixel scan, which is mainly achieved by moving detector or optic axis. Moving optic axis is performed by rotating or vibrating one optical component of the imaging lens. However, the machining requirements of the optical component for sub-pixel scan are rigorous or sometimes even impossible. A sur-pixel scan method is proposed, it can effectively restrain frequency aliasing while having lower machining requirements compared to sub-pixel scan. In this paper, the effect of sur-pixel scan on image frequency aliasing is theoretically analysed, its engineering superiority is mathematically calculated in visible spectrum and infrared spectrum. Simulations and experiments are performed to validate the effectiveness of sur-pixel scan method.