Gesture-Controlled Robotic Arm Utilizing OpenCV

摘要

In this paper, a low-cost, 3D printed robotic arm that uses a system of gesture recognition controlled via computer vision is presented. The developed approach of Human-Computer Interaction (HCI) employs a single commonly available USB2.0 High-Definition camera to capture hand movements and gestures, implementing the OpenCV library. This enables the translation of those movements and gestures into commands that can be transferred using internet protocols to a Raspberry Pi; then interfaces with the robotic arm. Using computer vision as a method of HCI reduces components and overall cost. The Raspberry Pi uses individual commands to control the robotic arm's servo motors while monitoring multiple pressure sensors and limit switches to ensure the robotic arm operates within a certain limit of load. This prevents the arm from overextending itself while protecting the integrity of items within its grasp.