Bearing estimation of screams using a volumetric microphone array mounted on a UAV
Detecting Human Screams with UAVs During Disasters
During disasters, such as earthquakes, fires, or collapsed buildings, every minute counts to find possible victims alive. The use of Unmanned Aerial Vehicles (UAVs) for Search and Rescue (SAR) purposes is rapidly increasing due to their ability to quickly survey large areas, that may be dangerous or difficult to reach, and provide situational awareness to first responders. Currently, UAVs use sensors to visually detect people when flying over disaster zones. However, in many cases, potential victims are not visible to these sensors. Victims could possibly be trapped in rubble or not visible due to opaque smoke.
Therefore, to help saving the lives of these victims, the Fraunhofer Institute for Communication, Information Processing, and Ergonomics (FKIE) in Germany is developing acoustic technology that can be attached to drones to determine the compass bearing of certain sounds, including human screams, during disasters. The acoustic technology consists of a special volumetric array of MEMS microphones, called “Crow’s Nest Array” (CNA), combined with advanced array processing techniques and artificial intelligence methods to achieve filtering, detection, bearing and source localization of specific audio events. In this forum, the CNA acoustic system will be presented, and its performance evaluated in various open field experiments, will be exposed.