An unmanned aerial vehicle target azimuth angle calculation method based on infrared and visible light fusion images
By fusing infrared and visible light images and transforming coordinate systems, the problems of low accuracy and poor robustness in calculating the target azimuth angle of small UAVs have been solved, achieving high-precision and stable target detection, especially with stronger anti-interference capabilities in complex environments.
Patent Information
- Authority / Receiving Office
- CN Β· China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2026-01-26
- Publication Date
- 2026-06-19
AI Technical Summary
Small fixed-wing UAVs, due to the lack of gimbal design and low-cost inertial measurement units, have large attitude measurement errors, resulting in low accuracy and poor robustness in target azimuth angle calculation, making it difficult to maintain stable target detection capabilities, especially in complex environments.
Infrared and visible light image fusion technology is adopted, and image fusion is performed through an end-to-end deep neural network. The target recognition and localization are combined with the YOLOv8 model. The coordinate system transformation is performed using the camera intrinsic and installation extrinsic matrix. Finally, the target azimuth angle is calculated, and the navigation transformation is performed using the NED coordinate system to achieve high-precision azimuth angle calculation.
It significantly improves the reliability and confidence of target detection, reduces the uncertainty of azimuth angle calculation, and enhances robustness to platform attitude disturbances. Through Monte Carlo simulation verification, the fused image shows higher detection consistency and lower statistical uncertainty in complex environments.
Smart Images

Figure CN122244143A_ABST