A2.3 - Automated Camera Tracking Software
- Event
- ETTC 2026 - European Test and Telemetry Conference
2026-06-09 - 2026-06-11
Nuremberg - Chapter
- Imaging & Video
- Author(s)
- F. M. Yildiz, M. K. Arpacioglu - Turkish Aerospace Inc., Ankara (Turkey)
- Pages
- 35 - 41
- DOI
- 10.5162/ettc2026/A2.3
- Price
- free
Abstract
Flight testing, ground-based cameras are used to provide an independent visual reference for validating aircraft motion, system behavior, and test execution in real time. Ground-based camera tracking of airborne platforms is traditionally achieved using image processing techniques, specialized PTZ hardware, or computationally intensive visual tracking algorithms. While effective, such approaches introduce increased system complexity, latency, and cost. In many operational and flight-test scenarios, however, the precise position of the aircraft is already available through real-time GPS telemetry. This study presents a geometry-driven camera tracking system that deliberately avoids image processing and instead relies solely on GPS data to control camera orientation. The proposed system computes realtime pan and tilt commands by combining the aircraft’s GPS position with the known geodetic location of each camera. The architecture supports simultaneous tracking by multiple cameras, where each camera independently calculates its own azimuth and elevation angles based on its fixed position. Camera control is performed using the ONVIF protocol, enabling compatibility with a wide range of standard commercial IP cameras without requiring specialized hardware. In addition to directional control, the system integrates automatic zoom and focus adjustment to maintain target visibility across varying distances, further enhancing tracking robustness without visual feedback. The resulting solution provides deterministic behavior, low computational load, and high scalability, making it well suited for flight test environments, temporary test ranges, and cost-sensitive surveillance applications. This study demonstrates that camera tracking problems can be effectively addressed through sensorbased geometric prediction rather than visual perception, offering a practical and efficient alternative for telemetry-driven observation systems.