Ball-Tracking-using-ROS

Personal Project

Ball Tracking using ROS and OpenCV:

Problem: Ball Tracking :The goal of this project was to write a an alogrithm for ball detection and tracking using Robotic Operating System(ROS) and OpenCV library.

Initially, the project was offered in one of the online MOOC course offered by Prof.Anis Koubaa.

Approach:

• The camera output is converted into HSV format.
• Color Filtering is done.
• Edge detection using Color Filtering.
• Once Edge detected, draw contour around it.
• Process the contour like calculating the area,perimeter of the contour.
• Disadvantage: It is not robust to lighting condition because of color filtering.

How to run:

• Download all the files in one folder and name in topic03_opencv
• Copy and paste the folder in your catkin workspace. (e.g ..\catkin_ws\src\)
• Once completed, go into catkin workspace. (..\catkin_ws\)
• Open the terminal in the current directory(..\catkin_ws\) and compile the file using:

catkin_ws

• Once done, open another terminal and run roscore
• To run the USB_CAM, run:

rosrun usb_cam usb_cam_node _pixel_format:= yuyv

• Then run the tennis_ball_usb_cam_tracker.py file
• To run:

rosrun topic03_opencv tennis_ball_usb_cam_tracker.py

Requirements:

• ROS MELODIC.
• UBUNTU 18.04 LTS.
• OPENCV 3.4.2 or later version.
• CVBRIDGE
• USB_CAM ROS PACKAGE.