I-SEE

The system uses computer‑vision and deep‑learning models to detect unmanned aerial vehicles in real time, providing 360° coverage with reported accuracy above 95 %. Video streams from cameras or network sources are processed on‑device, allowing edge‑only operation without cloud connectivity. Detected objects are tracked, classified by type and threat level, and their position, speed and trajectory are estimated with geolocation data.

Users can configure geofences and receive immediate alerts through visual, audible or messenger channels such as Telegram. The interface shows an interactive map with target markers, status indicators and historical tracks for up to ten objects. Video frames are enhanced with dynamic brightness correction and noise filtering before analysis.

In addition to detection and monitoring, the platform integrates with electronic‑warfare or kinetic countermeasures for active counter‑UAV response. It is positioned as a Ukrainian‑developed, experimental utility for continuous, 24/7 protection of airspace against drones.