Current outdoor AR/VR experiences suffer from fundamental localization limitations where GPS accuracy (3-5 meter precision) is insufficient for precise augmented reality content anchoring, while visual-inertial SLAM systems fail in GPS-denied environments or lack persistent spatial memory across sessions. Existing solutions operate as isolated single-device experiences without leveraging smartphone sensors and VR headset tracking in coordinated multi-device frameworks. This thesis presents a comprehensive framework for precision outdoor extended reality (XR) that addresses these limitations through multi-device sensor fusion, advanced localization algorithms, and persistent beacon-based infrastructure. The proposed system combines VR headset tracking and smartphone GPS/IMU to achieve sub-meter accuracy for outdoor AR content anchoring. The framework enables global-scale deployment of location-anchored augmented reality content with community-driven content creation and cross-session spatial memory, establishing a distributed system conceptually similar to the World Wide Web for georeferenced 3D information.
Le attuali esperienze AR/VR outdoor soffrono di fondamentali limitazioni di localizzazione dove la precisione GPS (3-5 metri) è insufficiente per l’ancoraggio preciso dei contenuti di realtà aumentata, mentre i sistemi SLAM visual-inertial falliscono in ambienti privi di GPS o mancano di memoria spaziale persistente tra sessioni diverse. Le soluzioni esistenti operano come esperienze isolate a singolo dispositivo senza sfruttare sensori di smartphone e tracciamento di visori VR in framework multi-dispositivo coordinati. Questa tesi presenta un framework completo per la realtà estesa (XR) outdoor di precisione che affronta queste limitazioni attraverso fusione sensoriale multi-dispositivo, algoritmi di localizzazione avanzati e infrastruttura persistente basata su beacon. Il sistema proposto combina tracciamento di visori VR e GPS/IMU di smartphone per raggiungere una precisione sub-metrica per l’ancoraggio di contenuti AR outdoor. Il framework abilita il deployment su scala globale di contenuti di realtà aumentata ancorati alla posizione con creazione di contenuti guidata dalla community e memoria spaziale tra sessioni, stabilendo un sistema distribuito concettualmente simile al World Wide Web per informazioni 3D georeferenziate.
Precision outdoor XR framework: multi-device GPS-vision localization with persistent beacon-based augmented reality
ZAMBETTI, MATTIA
2025/2026
Abstract
Current outdoor AR/VR experiences suffer from fundamental localization limitations where GPS accuracy (3-5 meter precision) is insufficient for precise augmented reality content anchoring, while visual-inertial SLAM systems fail in GPS-denied environments or lack persistent spatial memory across sessions. Existing solutions operate as isolated single-device experiences without leveraging smartphone sensors and VR headset tracking in coordinated multi-device frameworks. This thesis presents a comprehensive framework for precision outdoor extended reality (XR) that addresses these limitations through multi-device sensor fusion, advanced localization algorithms, and persistent beacon-based infrastructure. The proposed system combines VR headset tracking and smartphone GPS/IMU to achieve sub-meter accuracy for outdoor AR content anchoring. The framework enables global-scale deployment of location-anchored augmented reality content with community-driven content creation and cross-session spatial memory, establishing a distributed system conceptually similar to the World Wide Web for georeferenced 3D information.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Zambetti_Tesi.pdf
accessibile in internet per tutti
Descrizione: Tesi
Dimensione
2.15 MB
Formato
Adobe PDF
|
2.15 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Zambetti_ExecutiveSummary.pdf
accessibile in internet per tutti
Descrizione: Executive Summary
Dimensione
326.27 kB
Formato
Adobe PDF
|
326.27 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/252996