| Popis: |
This paper introduces the development of a system for a gunfire sound classification using a smartphone. The sound is detected using microphones connected to an embedded system to further process and transform it into a digital format. Artificial Neural Network is adopted as a recognition process for classifying the gunfire sound in the interested sound group. It is found that 6 different types of gunfire sound can be classified correctly with the proposed system. Also, the system can differentiate other sounds from the interested gunfire sounds. This proposed classifying system is expected to help analyze and find the source of the gun, in time, in the patrol area even better. |
