Measuring Quality of Experience
using Augmented Reality in Training Systems

TFE4580
Electronic Systems Design and Innovation, Specialization Project

Description • Installation • More Info • Contributors

This repository contains the specialization project done by Peter Remøy Paulsen, the fall of 2020, as a part of his masters degree in electronical engineering.

We took a deeper look into realism in AR, and tried to figure out what the level of occlusion has to say for the Quailty of Experience for the user in an augmented reality system.

Complete report can be found here.

Description

To test if occlusion can have anything to say, we built an augmented reality app using Unity and ARCore.

In the app the participants exposed of either environment 1 or environment 2

Environment	Description
Environment 1	Occlusion algorithm component from ARCores API enabled, so that the environment is able to capture depth data and occlude virtual objects in the real world
Environment 2	Occlusion algorithm component disabled. No dapth data captured. Virtual objects will float above real objects.

In action:

Environment 1	Environment 2

Installation

Pre-compiled APK

The pre-compiled APK can be found in the releases.

Compile on your own

You can compile the project on you own in Unity. Download the project and import the project to you Unity installation

Git Clone

HTTPS:

git clone https://github.com/petrepa/TFE4580.git

or SSH:

git clone [email protected]:petrepa/TFE4580.git

More info

Can be found in the Wiki

Contributors

• Researcher and Developer: Peter Remøy Paulsen 👨‍🎓
• Supervisor: Andrew Perkis 👨‍🏫
  • Co-supervisor: Shafaq Irshad 👩‍🏫

---

Further documentation needed for:

• Experimental Application Overview
  • With GIFs demonstrating anything
• TL;DR of report
  • But should link the entire report
• Data analysis
  • The Jupyter Notebook
• Casual notes
  • The wiki