Estimating human locomotion in GPS-denied environments: pedestrian dead-reckoning, direction-change intention and trajectory reconstruction. From body-worn inertial sensors, without external positioning infrastructure.
Objectives:
Pedestrian dead-reckoning and waist-worn inertial navigation
Accelerometry-based distance, speed and slope estimation
Step segmentation and stride-length modeling
Real-time short-term pedestrian trajectory prediction
Intention and turn anticipation from body kinematics