Research on breath detection algorithm of UAV under noise interference

Unmanned Aerial Vehicle (UAV), as an important aid for rescue work in complex environments, has the advantages of rapid deployment, autonomous patrol, and large-scale search and positioning, and can check vital signs without contact. However, due to UAV disturbance and difficult removal of ambient light noise, the error of breath extraction is large. In this study, an anti-noise breath detection algorithm was proposed. The optimal ROI region of the face was selected to extract the regions of interest in breathing to reduce the interference of face and environmental noise. The main components of breath were extracted by variational mode decomposition based on PSO optimization, which enhanced the anti-disturbance ability of the UAV. Experimental results show that the proposed algorithm has good accuracy and can effectively reduce noise interference.