Evaluation of different localization techniques for GNSS (global navigation satellite system) denied environments

At the slopes of the mountains, near tall buildings, in underground mining or even in agricultural scenarios, GNSS (mostly called GPS) signal is not always available or it is not robust enough. This project is aimed at implementing different localization techniques (i.e., dead reckoning, visual odometry, inertial odometry, SLAM -simultaneous localization and mapping- and point cloud strategies) and to compare them in terms of localzation errors, consistency and performance, to finally determine wich is the most suitable technique that could replace GPS signal when lost. To this end, it is expected to work with existing databases. No field experimentation is required.

Required skills: python/MatLab, statistics, signal processing.

Supervisor name: 
Fernando Auat Cheein
Supervisor email addresses: 
f.auat@hw.ac.uk