Supervisor

Suitable for

Abstract

Elements such as mirrors and glass still present a major difficulty for simultaneous localisation and mapping (SLAM) and 3D reconstruction algorithms [1]. This project will address this challenge by developing a method for accurately identifying mirrors in a scene and accounting for their unique properties in the reconstruction process. The goal will be to improve the fidelity and accuracy of the 3D models. The work will involve aspects of computer vision, geometric modelling, and possibly ray tracing techniques.

[1] https://www.thomaswhelan.ie/Whelan18siggraph.pdf

Pre-requisites: Suitable for those who have taken a course in machine learning or computer graphics