Geometric separation of superimposed images

This paper discusses a technique to separate two superimposed images where there is a geometric relationship between the two images. This technique would allow a single camera to capture two images simultaneously with differing fields of view. One image is a wide field-of-view (wFoV) image, and the other is a narrow field-of-view (nFoV) image corresponding to the center of the wFoV image. This is proposed as a means to generate situational awareness and object identification information from a single imaging sensor, with applications for surveillance onboard uninhabited aerial vehicles (UAVs). In this paper the results of the geometric separation technique are temporally averaged and demonstrated to be robust to image registration errors.