Geometric separation of superimposed images with varying fields-of-view

Using an imaging system with a wide field of view increases situational awareness (peripheral vision) at the expense of reduced image resolution. Ideally, an imager should provide high-resolution imagery around a specific region (or regions) of interest and a wide field of view. This paper proposes a particular approach to address these competing requirements: an image where a narrow field of view and a wide field of view are superposed on a single image frame and then subsequently separated in post-processing. The work introduces a recursive algorithm that can be used for the blind separation of two images from a single superposition image, using only prior knowledge of the geometry of the superposition.