Based on the 64-channel microphone array, the UAV acoustic model was established, and a UAV sound source localization method with improved MUSIC algorithm was proposed. After indoor and outdoor tests, the effects of different elevation directions and environmental noise on the positioning effect of UAV were studied. The experimental results show that in the indoor test, the positioning success rate of the MUSIC algorithm reaches 84.70%, which is 10.28% higher than that of the original algorithm, the anti-noise stability reaches 83.76%, and the positioning speed is 26.74  and the positioning success rate of the MUSIC algorithm reaches more than 93.72%, the anti-noise performance reaches more than 90.47%, and the positioning speed is more than 28.16 fps, which verifies the feasibility of the improved algorithm. Combined with the Hidden Markov model (HMM) scanning model, the accurate scanning of high-altitude intrusion UAVs is realized.