Monocular Depth Estimation Using Rhaznoff Maneuver

Accurate range knowledge is essential for electro optical guidance because it fixes the size of the tracking gate, scales lead angle commands in proportional navigation laws, and sets impact and fuze timing. Traditional seekers obtain this information with stereo cameras or laser range finders, but those sensors increase cost, mass, power consumption, and integration complexity. These penalties can be unacceptable for disposable munitions considering drones and missiles. The paper presents a vision and kinematic framework that determines target depth using a single monocular camera under certain but feasible conditions, removing the need for any extra range sensor. The algorithm used in this work utilizes only basic vector operations, finishes within similar time sufficient for basic vector operations on a standard personal computer, and works without prior calibration, except for the camera intrinsics. Simulations that include tail chase scenarios yielded errors under 3 percent from different ranges which is sufficient for correct lead commands and safe warhead arming. The same approach can also give low-cost depth estimates for indoor mobile robots and other compact autonomous platforms showing its further application domains.