Low-Light Micro-PIV Images of Supercritical CO2 Flows with Paired Ground Truth

17.9 GB of experimentally acquired low-light image sequences show fluorescent particles in carbon dioxide flows under high-pressure, transcritical conditions. Each optically distorted image sequence is paired with a speckle-free ground-truth image of an identical particle pattern, accompanied by thermodynamic metadata in CSV files. The dataset, authored by Norbert Krauss and last updated in April 2026, is designed for algorithm development and validation.

Use Cases

• Benchmarking image reconstruction algorithms based on paired corrupted and ground-truth image sequences.
• Developing data-driven denoising models for µPIV data based on low-light images with optical distortions.
• Validating conventional image processing techniques under realistic experimental conditions described in the dataset.
• Analyzing particle behavior in supercritical fluid flows using the provided thermodynamic metadata.

Strengths

• 17.9 GB dataset provides substantial volume for training and testing.
• Each corrupted image sequence is paired with a speckle-free ground-truth for direct algorithm validation.
• Thermodynamic metadata in CSV files links each image to precise experimental conditions.
• Includes reference image data for independent benchmarking of particle properties.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain tasks.

Provenance

Source

figshare

Collection Method

Experimentally acquired using a high-speed CCD camera imaging fluorescent particles through a microchannel.

Time Range

null

Freshness

Last updated 2026-04-09 10:05:08

Geography

null