Gesture recognition for Human-Robot interaction

Abstract

The goal of this project was to implement a real-time system to recognize the hand’s gesture for Human-Robot Interaction (HRI). In this project, the recognition consists of 2 steps: Tracking and Recognition. The hand tracking algorithm involved color-based skin segmentation, difference with face region and the Recognition accomplished by applying Neural Networks.

A software-based algorithm was examined in MATLAB to evaluate its performance and verify its effectiveness. However, the speed of processing in MATLAB is quietly slow, so the result would have a few limitations.

Although the system can be improved to obtain better results, overall the project was success as it enabled any inputted hand to be accurately recognized.