Event Ellipsometer: Event-based Mueller-Matrix Video Imaging

Ryota Maeda, Yunseong Moon, Seung-Hwan Baek

[Project Page] [Paper]

Software Environment Setup

Python Environment

To set up Python environment, install the libraries specified in pyproject.toml. If you are Rye user, you can run rye sync to set up the environment.

C++ Extension

We developed a C++ extension for the event data processing to accelerate the computation The extension needs to be built manually by running the build_cpp.py. When building the extension, Eigen and OpenMP are required. We tested this extension on Windows.

Event Camera and Data

Metavision SDK is necessary for acquiring data from event cameras and handling data I/O. Please ensure the SDK is installed and can be imported in Python using the command import metavision_core.

Hardware Development

We developed a hardware prototype using Arduino IDE. To calibrate the contrast threshold, we utilized WaveForms to operate the Analog Discovery.

Running the Code

python scripts/reconstruct_mm.py --filename_raw recordings\recording_2024-11-07_15-55-17_small.npz

The file recording_2024-11-07_15-55-17_small.npz is a sample event data converted from a raw data file.

The full dataset will be available soon.

BibTeX

@article{maeda2024event,
    title={Event Ellipsometer: Event-based Mueller-Matrix Video Imaging},
    author={Maeda, Ryota and Moon, Yunseong and Baek, Seung-Hwan},
    journal={arXiv preprint arXiv:2411.17313},
    year={2024}
}