Computer Vision based hand mouse cursor system

Instructor: Scott MacKenzie

Design and implementation of a hand tracking and hand gesture classification system used to demonstrate it’s capability as an adjunct to a computer mouse. Link. Used several technologies for its implementation in Matlab:

  • Image Processing Algorithms: Color segmentation, Background subtraction, Histogram of Gradient (HOG) feature extraction
  • Hand Tracking Algorithm: Kalman Filter
  • Real time Gesture Classification using Matlab’s Neural Network Toolbox
Applying Computer Vision and Natural Language Processing to real world problems.