Automatic keystone correction for smart projectors with embedded camera

Abstract

We propose embedded algorithms for enabling a smart self-adjusting projector that is capable of fully automatic keystone distortion detection and correction, plus zooming and centering the image to fit a screen. The projector has an embedded camera and it automatically detects the screen and a specific pattern during self-adjustment. Proposed algorithms meet severe computational requirements of a portable projector and they exploit the physical constraints in a projector-(embedded) camera pair in a novel way in order to not compromise image quality under different usage conditions. We will demonstrate our projector system during the conference.