Welcome to UCF's contributions to the OGC UDTIP Pilot Project, focusing on standardized pipelines for geospatial data integration. This repository documents my work for the D101 deliverable, emphasizing GeoPose-enabled Camera Imagery Interoperability, GeoAI, and Machine Learning within Urban Digital Twin systems.
- D101 Deliverable: OGC UDTIP D101 Artifact
- Related Paper: GeoPose-enabled Camera Imagery Interoperability with Geo-AI in Urban Digital Twins
Our work advances open, interoperable urban digital twins by:
- GeoPose-enabled Camera Imagery Interoperability
- TrainDML with GeoAI & Machine Learning
- Urban Digital Twin Systems Integration
- D101: Camera Imagery Interoperability (GeoPose, standardization, data pipelines)
- D102: TrainDML-AI standard & ML for GeoAI applications
Our approach leverages Inertial Navigation Systems (INS) for high-precision pose estimation, outperforming traditional IMU and AHRS solutions.
We developed a robust GeoPose script for converting INS data to standardized GeoPose format, ensuring interoperability.
Sample output of a standardized GeoPose Sequence Regular Series in JSON format:
Synchronizing real-world heterogeneous sensor data with GIS information is essential for accurate GeoPose generation and urban digital twin fidelity.
A comparative analysis of datasets for urban digital twin applications, including our custom dataset.
Our D102 deliverable demonstrates the use of the TrainDML-AI standard and machine learning for GeoAI applications on captured data.
This work is part of the OGC Urban Digital Twin Interoperability Pilot, with contributions from UCF and the broader OGC community.
For questions or collaboration, please open an issue or contact the contributors through the OGC portal.