The problem of wheelchairs is becoming more serious now as many disabled people need wheelchairs for mobility. The varied condition of the road surface in Indonesia can cause discomfort and the risk of accidents for wheelchair users. Wheelchair users currently have to control speed manually, but the Image Recognition feature for automating electric motor speed can improve comfort and reduce the risk of accidents. Therefore, the proposed solution is to develop a smart wheelchair system that uses the MobileNetV2 Image Recognition method to control electric motor speed according to the condition of the road surface. The developed smart wheelchair system is able to identify the type of road surface traveled using the MobileNetV2 method and adjust the wheel's speed according to the needs. The results of the testing of this system in the form of percentage of prediction accuracy percentage of 97% with a computing time of 0.24-0.25 seconds.
Copyrights © 2023