Target geolocation system and method based on drone imagery
By using a target geolocation system based on UAV imagery and employing a 2.5D reference map and PnP ray projection algorithm, the problem of limited accuracy in UAV target geolocation was solved, achieving high-precision and robust geolocation and promoting the standardization of the technology.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NAT UNIV OF DEFENSE TECH
- Filing Date
- 2026-02-06
- Publication Date
- 2026-05-29
AI Technical Summary
Existing UAV target geolocation methods are limited by UAV attitude estimation errors and the availability of accurate altitude prior data, resulting in limited positioning accuracy and difficulty in achieving efficient and accurate target geolocation in complex environments.
A target geolocation system based on UAV imagery is adopted, including a data acquisition module, an end-to-end geolocation module, and a benchmark testing and evaluation module. It utilizes a 2.5D reference map, feature matching, and a PnP-based ray projection algorithm, combined with a RANSAC attitude estimation strategy, to achieve high-precision 3D geolocation.
It improves geolocation accuracy and system robustness, adapts to multiple scenarios and environmental conditions, provides objective evaluation standards, and promotes the standardized development of target geolocation technology.
Smart Images

Figure CN122108064A_ABST