Home

Design of a low-cost, vision-based motion capture system for estimating the pose of rigid-body platforms

Vision-based motion capture system
Vision-based motion capture system



Description

This project aims to design and create a motion capture system that utilises multiple cameras to estimate the position and pose of a moving object. The cameras are used to capture pixel coordinate data of certain features of the robot, then transmit the data to a base station which runs an EKF algorithm to estimate position, velocity, orientation and angular velocity.

Key skills/interests:

Mathematical modelling, machine vision, state estimation

Expected outputs:

1x MSc (Eng), 1x Journal/Conference paper in a leading publication.

Supervisor:

Arnold Pretorius