Back to Projects

Vision based Motion Detection and Tracking

This was a course project for Advanced Vision (MSc. Artificial Intelligence, University of Edinburgh). The goal was to build a system for detecting motion in an indoor environment and tracking detected objects based on color distribution using a stationary color camera.

Motion Detection Example

Highlights Video

Overview

A background image served as a reference. Motion was detected via simple pixel subtraction between the background and subsequent frames. Detected moving objects were classified and assigned IDs based on color information.

In later frames, objects were tracked by matching their color distribution (using Bhattacharyya distance) to objects detected in the previous frame, maintaining consistent IDs.

Technical Approach

  • Background subtraction for motion detection
  • Color histogram-based object representation
  • Bhattacharyya distance for histogram matching
  • Persistent object ID assignment across frames

Course Project

Developed as part of the Advanced Vision course during the MSc. in Artificial Intelligence program at the University of Edinburgh.