Abstract
The detection of low-altitude slow-small (LSS) targets, such as drones, is challenged by their small radar cross-section (RCS) and low signal-to-clutter ratio (SCR), resulting in short effective range and susceptibility to background clutter in complex environments. To overcome the limitations of conventional radar and electro-optical methods, this paper proposes a novel detection theory based on broadband spectral modulation imaging (BSMI). We analyze the recognition accuracy for drone targets across different zenith angles and detection ranges through numerical simulations. A snapshot-based BSMI detection system was designed and implemented, with experiments conducted under consistent conditions for validation. Results demonstrate that the system achieves over 90% classification accuracy, confirming the theory's effectiveness. This study significantly enhances detection probability and suppresses false alarms for low-altitude drones, providing a viable technical solution for monitoring unauthorized aerial activities.