The aim of this project is to assist people with hearing disabilities by identifying, localizing, and classifying alerting sounds, such as car horns and ambulance sounds. The detection of the surrounding sounds is performed using a custom-made head hat, in which four microphones are attached. The detected sounds are sent to a microcontroller that continuously analyzes the detected signals to identify alerting sounds. In addition, the microcontroller uses cross-correlation analysis to localize the source of the alerting sounds. Upon the detection of an alerting sound, the microcontroller alerts the person about the alerting sounds by running vibrating motors attached to the head hat to generate mechanical vibrations in region that matches the direction of the alerting sound. Moreover, the microcontroller promptly sends a trigger via a Bluetooth channel to a Windows smartphone, which uses its microphone to record the alerting sounds. The smartphone analyzes the record signal using cepstrum analysis to classify its type. The type of the alerting sound is visually communicated to the person via the screen of the smartphone.
Done by:
Students: Mahmoud Al-Ashi, Faris Abawi
Dr. Mohammad Daoud, Dr. Ala’ Khalifeh