An open-access dataset of tomographic X-ray data of a walnut, created by K. Hämäläinen of the University of Helsinki. It includes sinograms at multiple resolutions, corresponding measurement matrices, and a high-resolution ground truth reconstruction from 1200 projections. Documentation is available in a referenced arXiv paper.
Use Cases
- Benchmarking tomographic reconstruction algorithms based on the provided sinograms and measurement matrices.
- Developing machine learning models for image reconstruction using the provided ground truth FBP reconstruction.
- Studying the effects of down-sampling and projection count on reconstruction quality based on the multiple resolution sinograms.
Strengths
- Includes a high-resolution ground truth reconstruction computed from 1200 projections.
- Provides sinograms at three different resolutions derived from a measured 120-projection fan-beam sinogram.
- Original measured sinogram is included in its original form and resolution.
Limitations
- Row count and dataset size are unknown, which may limit suitability assessment.
- Column-level documentation is absent; field semantics must be inferred after download.
- Last update date is unknown; freshness unverified.
Provenance
- Source
- University of Helsinki
- Collection Method
- Measured X-ray tomographic data from a walnut, processed via down-sampling and logarithmic transformation.