COMP3204 Computer Vision

Coursework 1: The OpenIMAJ Tutorial


Maintained by Professor Mark Nixon and Dr Jonathon Hare.

Brief

Due date: Friday 26th November, 16:00.
Link to the tutorial: html or pdf
Handin: 2122/COMP3204/2/
Required files: code.zip
Credit: 10% of overall module mark

Overview

OpenIMAJ is a set of open-source libraries and tools for multimedia content analysis and content generation that has been written for research projects at Southampton over many years. OpenIMAJ is very broad and contains everything from state-of-the-art computer vision and advanced data clustering, through to software that performs analysis on the content, layout and structure of webpages. You will need to use OpenIMAJ for all the COMP3204 courseworks.

For this coursework, we want you to work through a portion of the OpenIMAJ tutorial in order to get a grasp of the underlying functionality and gain some experience in practical image processing and computer vision. Even though this is the first coursework to be set, the deadline is deliberately late in order to allow you to work through the tutorial in line with the underlying computer vision techniques as they are introduced in the lectures.

You can consider this coursework to be like a set of labs in that the only requirement to get credit is that you provide us with evidence (i.e. code) that you have completed the relevant sections of the tutorial.

Details

In order to complete the second coursework (which will be announced during Mark’s lectures) you will need to have worked through Chapter 1 and Chapter 2. During Jon’s part of the course the handout’s accompanying the lectures will highlight the relevant chapter of the tutorial that demonstrates the topic of that lecture; these chapters of the tutorial are a prerequisite to successfully completing the third coursework.

Summary of required tutorial chapters:

Before starting the second coursework (i.e. do it as soon as possible!):

Before starting the third coursework (you can do these to follow along with the lectures):

What to hand in

You need to submit via ECS Handin a zipped copy of your code. Specifically, you should submit a Maven project and accompanying Java source code which show that you have worked through the above chapters of the tutorial. Make sure you include evidence of have attempting the exercises associated with each tutorial - if an exercise asks a question or for reflection, make sure you leave this as a comment in the code. Please don’t include the Maven ‘target’ directory in your submission. We would recommend you structure your code so that each chapter of the tutorial is in a separate Java package (i.e. uk.ac.soton.ecs.<userid>.ch1, uk.ac.soton.ecs.<userid>.ch2, etc.).

Marking and feedback

Marks will be awarded for:

  • Successful completion of the task.
  • Evidence of understanding.
  • Good use of the OpenIMAJ library.
  • Well structured and commented code.
  • Evidence of professionalism.

Standard ECS late submission penalties apply.

Individual feedback will be given covering the above points.

Questions

If you have any problems/questions then email or speak to Jon in his office.

Copyright ©2021 The University of Southampton. All rights reserved.