UAV propeller frequency and angle analysis using lidar and sliding windows

While extensive research has been conducted on UAV detection using various modalities, limited work has addressed characterizing the frequency and angle of rotating propellers. Identifying frequencies could help solve the problem of classifying birds vs drones. This study aims to answer the question: "How accurately can the frequencies of a stationary UAV rotor be identified at various angles in the spectral analysis of lidar pulses?" We propose a lidar-based solution to detect both frequency and angle through spectral analysis. Using a sliding window to identify the propeller signal in a spectrogram, we found that frequency measurement is highly dependent on the beam’s position on the rotor and the angle between them. Our results indicate that frequency measurements are more accurate for beams positioned directly on the rotor blade and when the angle between the rotor and the beam approaches 90 degrees. However, predicting the angle of the propeller proved to be inconclusive.