Description
Students will gain knowledge about robot real-time pose estimation and mapping, with an emphasis on the use of vision as a primary sensor for mapping the environment. The module will provide students with an understanding and practical experience of how to combine information from satellite navigation and motion sensing systems, recover geometry from optical sensors and creating an environment map which a robot can use for navigation and motion planning.
Ìý
To navigate safely, robots need the ability to localize themselves autonomously using their onboard sensors. Potential tasks include the automatic 3D reconstruction of buildings, inspection and surveillance.
Ìý
Aims:
ÌýThe aims of this module are to:
- Support students to design and optimize robot vision and navigation systems that can operate reliably and accurately in real-world environments.
- Develop students’ knowledge about robot real-time pose estimation and mapping, with an emphasis on the use of vision as a primary sensor for mapping the environment.
- Develop students’ understanding and practical experience of how to combine information from satellite navigation and motion sensing systems, recover geometry from optical sensors.
- Develop students’ understanding and practical experience of creating an environment map which a robot can use for navigation and motion planning.
Intended learning outcomes:
On successful completion of the module, a student will be able to:
- Apply fundamental techniques used for real-time estimation in linear and nonlinear systems.
- Formulate algorithms to fuse data from satellite and motion sensing systems to estimate robot position.
- Formulate mapping and localisation problems in which robots construct sparse maps of their environment.
- Create 3D reconstructions of the environment using camera data.
- Programme with Matlab or Python or C++.
Indicative content:
The following are indicative of the topics the module will typically cover:
- Mathematical formulation of the SLAM problem.
- Graphical models and probability.
- Sparse SLAM algorithms (system - ORB-SLAM2.)
- Grid and volume-based algorithms (systems - Octomap.)
- Normalised Distributed Transformations.
- Implicit representations such as NeRFs.
Requisite conditions:
To be eligible to select this module as optional or elective, a student must be (1) registered on a programme and year of study for which it is formally available; (2) have completed an introductory module on machine learning.
Module deliveries for 2024/25 academic year
Last updated
This module description was last updated on 19th August 2024.
Ìý