Lucas-Kanade-Tracker

Overview

This project implements object tracking using Lucas-Kanade template tracking. Object tracking was implemented to track 3 things: a car, face of a human baby, and a human. 3 different dataset were used for each of them. You can find these dataset here.

The output videos can be seen here.

Dependencies

• Python3
• Python3-tk
• Python3 Libraries: Numpy, OpenCV-Python, Scipy

Install Dependencies

• Install Python3, Python3-tk, and the necessary libraries: (if not already installed)

sudo apt install python3 python3-tk
sudo apt install python3-pip
pip3 install numpy opencv-python scipy

• Check if your system successfully installed all the dependencies
• Open terminal using Ctrl+Alt+T and enter python3.
• The terminal should now present a new area represented by >>> to enter python commands
• Now use the following commands to check libraries: (Exit python window using Ctrl + Z if an error pops up while running the below commands)

import tkinter
import numpy
import cv2
import scipy

Run

• Download each of the dataset mentioned in the Overview Section.
• It is recommended that you save the dataset within outer-most directory-level of the project otherwise it will become too cumbersome for you to reference the correct location of the file.
• Using the terminal, clone this repository and go into the project directory, and run the main program:

https://github.com/urastogi885/lucas-kanade-tracker
cd lucas-kanade-tracker/Code
python3 main.py dataset dataset_location output_location select_roi

• If you have a compressed version of the project, extract it, go into project directory, open the terminal by right-clicking on an empty space, and type:

cd Code/
python3 main.py dataset dataset_location output_location select_roi

• For instance:

python3 main.py baby ../DragonBaby/img/ ../DragonBaby/output.avi 0

• Choose select_roi as "0" to use saved ROI points and "1" to select ROI region yourself.
• Use the following to define the dataset-parameter in the input arguments:
  • car - Car Dataset
  • bolt - Bolt Dataset
  • baby - Dragon Baby Dataset
• For further documentation on the input arguments, refer main.py