GNSS Positioning with Low Cost Devices

Outdoor quantitative analysis of human movement is constrained by the technologies currently available. Marker-based optical motion capture is laboratory-bound. Inertial measurement units are wearable but lack an absolute spatial reference and suffer from integration drift. Markerless video-based systems introduce camera geometry biases and operational constraints. None of these approaches simultaneously satisfies the requirements of outdoor, infrastructure-light, deployable motion capture. This thesis investigates whether wearable Global Navigation Satellite System (GNSS) receivers, combined with inertial sensing and a portable reference station, can bridge that gap. A conceptual framework identifies three functional roles for GNSS in human motion analysis and operationalises them into three integration architectures of increasing scope. The thesis validates this framework through three experimental contributions of correspondingly increasing complexity, with Architecture 2 tested rigorously, Architecture 1 evaluated incidentally through a single reference receiver, and Architecture 3 demonstrated as a proof of concept. A foot-mounted GNSS/IMU device achieved centimetre-level horizontal foot-trajectory accuracy at slow and normal walking speeds in a single-subject proof-of-concept study, and characterised the mechanism by which impulsive foot accelerations degrade carrier-phase tracking. A nine-subject outdoor campaign with a portable base station confirmed the degradation mechanism across subjects and yielded internally consistent stride parameters (r = 0.997 against a torso-derived walking speed reference), reflecting the near-identity relationship between stride length, cadence, and speed. A proof-of-concept multi-segment kinematic reconstruction with three GNSS/IMU devices and an OpenSim biomechanical model reproduced hip-flexion timing against a drone-based markerless reference (Pearson correlations between r = 0.752 and r = 0.926), with amplitude accuracy characterised separately through range-of-motion ratios. The thesis establishes the conceptual, methodological, experimental, and software foundations for outdoor GNSS-aided human motion capture, with an operating envelope appropriate for further development towards clinical rehabilitation assessment.