Abhishek Patil
← Back to Projects Hand Gesture Recognition

Hand Gesture Recognition

hand gesture recognitionOpenCvroboticsROS

Goal

Using a depth camera, OpenCV and ROS, develop an API to recognize number of fingers from 0 to 5.

Demo

Project Website

This project was integrated into another bigger project on TurtleBot navigation and can be viewed here: TurtleBot Navigation

Project Details

The project can be summarized in the following 5 steps:

  1. Extract hand region from raw depth data.
  2. Determine convex hull and convexity defects of extracted hand region.
  3. Detect no. of fingers using the convexity defects from (2).
  4. Create a custom image window to show the recognition in real-time.
  5. Integrate with ROS.

The code (integrated with ROS & the TurtleBot project) is available on GitHub: patilnabhi/tbotnav

Future Work

Improve the fingers recognition results by combining with skin-color segmentation and machine learning.

This project was completed as part of the MS in Robotics (MSR) program at Northwestern University.

References