EBIV - Event-based image velocimetry

Python and C++ code implementations for the analysis of event-camera recordings obtained from fluid flows.

Input: raw event data represented as time-surfaces

Output: Recovered velocity field

Updates

June 2024:

• added source code for pyEBIV Python interface including compilation instructions for Windows using Visual Studio commandline toolset (VS2022)
• sample raw data of water jet recorded with pulsed-EBIV including Python script showing pulsed event-data handling
• added PyBind11-based interface to library

Installation using PIP

Download and unpack the repository.
Open a Python console and navigate to the ./pyebiv directory.
Run the following PIP command:

pip install .

This assumes that PyBind11 and a suitable build system such as Visual Studio commandline tools are installed. The resulting interface will be copied to the ./site_packages directory of the Python distribution.

Publications

[Associated publications]

• "Event-based imaging velocimetry: an assessment of event-based cameras for the measurement of fluid flows"
  (Experiments in Fluids 63, Article number 101 (2022), DOI 10.1007/s00348-022-03441-6)
  [Open access] https://doi.org/10.1007/s00348-022-03441-6

• "Event-based imaging velocimetry using pulsed illumination"
  (Experiments in Fluids 64, Article number 98 (2023), DOI 10.1007/s00348-023-03641-8)
  [Open access] https://doi.org/10.1007/s00348-023-03641-8

@article{EBIV:2022,
    title = {Event-based imaging velocimetry: an assessment of event-based cameras for the measurement of fluid flows},
    author = {Willert, Christian and Klinner, Joachim},
    doi = {10.1007/s00348-022-03441-6},
    url = {https://doi.org/10.1007/s00348-022-03441-6},
    journal = {Exp. Fluids},
    volume = {63},
    pages = {101},
    year = {2022},
}

@article{PulsedEBIV:2023,
    title={Event-based imaging velocimetry using pulsed illumination},
    author = {Christian Willert},
    doi = {10.1007/s00348-023-03641-8},
    url = {https://doi.org/10.1007/s00348-023-03641-8},
    journal = {Exp. Fluids},
    volume={64},
    pages={98},
    year = {2023},	
}

Further reading

https://github.com/uzh-rpg/event-based_vision_resources

Acknowledgements

Part of the pyEBIV interface makes use of Prophesee's Metavision SDK, in particular portions of the raw-file decoder (metavision_evt3_raw_file_decoder.cpp), licensed under the Apache License, Version 2.0.