Control of a robotic arm through hand gestures using computer vision

This thesis presents a vision-based system for controlling a robotic arm through hand gestures, enabling intuitive and contactless human-robot interaction. The proposed system captures hand images using a standard RGB camera, processes them to detect hand landmarks, and classifies gestures using computer vision techniques. These recognized gestures are then mapped to control commands for the robotic arm. The research explores key topics such as 3D transformations, gesture recognition methods, and robotic kinematics, integrating state-of-the-art frameworks like MediaPipe and ROS. Experimental results demonstrate the system's effectiveness in real-time control scenarios, highlighting its potential for applications in assistive robotics, industrial automation, and sterile environments.