Table of Contents

Tracking

THIS MODULE IS NOT AVAILABLE SINCE IT IS INCOMPLETE.

Particle filter, Kalman Filter and face tracking

Module #1

Particle filter

  1. Study particle filter tutorial
  2. Watch tutorial video part 1
  3. Watch tutorial video part 2
  4. Study Matlab implementation

Module #2

Kalman filter

  1. Study Kalman filter tutorial
  2. Watch tutorial video part 1, 2 and 3.
  3. Study Matlab implementation.
  4. Write a Matlab script to study the performance compared with Particle filter (the same trajectory with different tracking trajectories).

Module #3

Kalman filter in 2D

  1. Study Object tracking in 2D tutorial
  2. Study Matlab implementation

Module #4

Face tracking

  1. Now you familiar with Kalman filter for tracking.
  2. Try to track human face in the video!
  3. You can use this short video, with the object bounding box as [320.8751 175.1776 103.5404 129.0504](top left width height) at the first frame. Track the center of the box and see how it works.
  4. You can also use built-in Matlab function to detect face every k frame. See how it helps and comment on what you observe.

/* Matlab. http://dlib.net/correlation_tracker.py.html */

ECE4580 Learning Modules