We design and implement an efficient architecture for geometric computation of the global position of an airborne video camera from images of known landmarks. A robust Hough transform-like method facilitated by a class of CORDIC-structured computations is implemented. It empowers aerial surveillance systems to navigate effectively when the GPS and inertial navigation sensors are out of order or GPS functionality is disrupted by jamming etc. Our architecture exploits parallelism, data correlation and uses pipelining with minimal double precision division for superior performance.