Omega-Prime: Data Model, Data Format and Python Library for Handling Ground Truth Traffic Data

Data Model, Format and Python Library for ground truth data containing information on dynamic objects, map and environmental factors optimized for representing urban traffic. The repository contains: Data Model and Specification see ./docs/omega_prime_specification.md 🌍 Data Model: What signals exist and how these are defined. 🧾 Data Format Specification: How to exchange and store those signals. Python Library 🔨 Create omega-prime files from many sources (see ./tutorial.ipynb): ASAM OSI GroundTruth trace (e.g., output of esmini) Table of moving object data (e.g., csv data) ASAM OpenDRIVE map LevelXData datasets through lxd-io extend yourself by subclassing DatasetConverter 🗺️ Map Association: Associate Object Location with Lanes from OpenDRIVE or OSI Maps (see tutorial_locator.ipynb) 📺 Plotting of data: interactive top view plots using altair ✅ Validation of data: check if your data conforms to the omega-prime specification (e.g., correct yaw) using pandera 📐 Interpolation of data: bring your data into a fixed frequency 📈 Metrics: compute interaction metrics like PET, TTC, THW (see tutorial_metrics.ipynb) 🚀 Fast Processing directly on DataFrames using polars, polars-st The data model and format utilize ASAM OpenDRIVE and ASAM Open-Simulation-Interface GroundTruth messages. omega-prime sets requirements on presence and quality of ASAM OSI GroundTruth messages and ASAM OpenDRIVE files and defines a file format for the exchange and storage of these. Omega-Prime is the successor of the OMEGAFormat. It has the benefit that its definition is directly based on the established standards ASAM OSI and ASAM OpenDRIVE and carries over the data quality requirements and the data tooling from OMEGAFormat. Therefore, it should be easier to incorporate omega-prime into existing workflows and tooling. To learn more about the example data read example_files/README.md. Example data was taken and created from esmini.