This solution detects the location of a sound source that corresponds to an azimuth using sound entered with three microphones. Estimating the azimuth value between sound source and microphones by deriving the difference of time at which sound is entered with microphones, it locates where a specific sound of voice was generated.
Through linkage with voice and image recognition technologies, it can be used in diverse applications, such as to eliminate CCTV blind spots or to provide a platform for conversation with intelligent robots.
Through linkage with voice and image recognition technologies, it can be used in diverse applications, such as to eliminate CCTV blind spots or to provide a platform for conversation with intelligent robots.
Serveillence
Using a real-time target sound detection algorithm, the target sound is detected in real-time and, concurrently, the location of sound source is detected.
Recognition rate with an allowable angle error by ±5˚ is guaranteed based on a distance of 5m from the home environment.
Recognition rate with an allowable angle error by ±5˚ is guaranteed based on a distance of 5m from the home environment.
- Azimuth mapping
- Increasing in clockwise direction with center point of channel 1-3 as 0 degree
