Gesture - Controlled Disability EV System

Abstract

Hand gesture recognition provides an intuitive and contactless method for human–machine interaction, especially in intelligent transportation systems. Traditional bike control mechanisms rely on physical interfaces that may limit accessibility and ease of operation. This research aims to design and implement a Hand Gesture Controlled Bike using computer vision techniques and an ESP32 microcontroller to enhance user interaction and control flexibility. The proposed system uses a camera to capture real-time hand gestures, which are processed using Python-based computer vision methods. Image acquisition and preprocessing are performed using the OpenCV library, while the MediaPipe framework is employed for accurate hand landmark detection and finger tracking. Recognized hand gestures are mapped to predefined bike control commands such as forward, backward, left turn, right turn, and stop. These commands are transmitted to the ESP32 through serial communication, where the microcontroller controls the motor driver to execute the corresponding bike movements. A graphical user interface is developed to display the live video feed and detected gestures, enabling real-time monitoring. Experimental results demonstrate that the system can accurately recognize gestures and reliably control bike movements with minimal delay. The proposed approach offers an efficient, low-cost, and user-friendly solution for gesture-based vehicle control and contributes to the development of smart transportation and assistive mobility systems.