Disability in Indonesia reaches 5% of the Indoensian population, including physically and visually impaired people. Due to the limited mobility experienced by people with disabilities, it is necessary to develop assistive devices that can helmp people with disabilities in their activities. Eleectric wheelchair has the advantages of being easy to control and maneuver, theri ability to go over slopes and ease to transport to realitively distant places. This study aims to develop a system that can be used to automatically detect room name plate on a smart wheelchair. In this system, the object of the room name plate captured by the camera will be the target of the automatic movement of the wheel chair. the captured camera object will be processed using the YOLOv5 and EasyOCR. this study can prove that the accuracy for the name plate detection is 60% and 100% accuracy for the character recognition. This allows the integration ability of the object detection and room character recognition systems in the autonomous system to trigger the PWM (Pulse WIth Modulation) value, which can be interpreted as if the system cannot recognize the character of the room name plate, resulting in the wheelchair continuing to move forward until it;s able to ding the desired room name.
Copyrights © 2023