Virtual Booth

Communication barriers between hearing and deaf community remain a significant challenge, as few in the hearing community can interpret sign language. With fingerspelling, the deaf community has the option to initiate the communication by spelling out their words to convey their message to the hearing community. Hence, this project presents a prototype of a fingerspelling recognition system that can translate the fingerspelling into the gestured alphabet. It uses sensor-based technology in detecting the fingerspelling performed by the deaf individual; and translate them into the alphabet for the hearing community to understand. The fingerspelling signals are collected by the flex sensors attached on a glove worn by the deaf user; while the translated alphabet was printed in the OLED display. An Arduino MEGA was used as the platform to process the signal into its detected alphabet. Initial testing of the fingerspelling recognition prototype resulted up to 76.15% accuracy, indicating that the sensor signal has potential in providing accurate information on the gestured fingerspelling. With this prototype, the communication between the hearing and deaf community can be made possible through the alphabets translated, without the need for the hearing community to know sign language.