Passive Wide-Area Three-Dimensional Imaging

Active 3D imaging methods, such as radar and lidar, often use large, heavy sensors that consume significant amounts of energy. Passive 3D imaging techniques, based on feature matching, are computationally expensive and constrained by the quality of the feature matching, making them less efficient. Currently, both these methodologies for 3D imaging pose challenges. Heavy-duty active 3D sensors reduce mobility and increase energy requirements. Meanwhile, passive 3D imaging techniques necessitate high-quality feature matching while also introducing computational bottlenecks. The drawbacks of both active and passive imaging motivate the development of a lightweight, energy-efficient, and computational cost-effective technology for 3D imaging.

Technology Description

The technology enhances 3D image formation from full-motion videos using a cost- and energy-effective method. Instead of relying on large, power-burning, active 3D imaging sensors, this technology involves registering and mapping video frames to the scene coordinates by leveraging the data from the platform's trajectory with respect to the scene. This method uses a mathematical relationship that involves the height map of the scene, projected angular velocity of the platform, and spatial gradient of the registered frames. Creatively circumventing the limitations posed by both active and passive 3D imaging techniques, this technology distinguishes itself through the use of full-motion video acquired from a moving platform. The primary differentiator lies in the real-time processing ability to produce height maps of the scene from a full-motion video and trajectory, proving both efficient and robust.