SEDDI DOME: Digitized Material Appearance at Micro and Meso Scales

A dataset of material samples digitized at two scales using a hemispherical light dome. The data includes raw captures and optimized parameters for geometric properties, anisotropic reflectance, and transmittance, released by author Elena Garces in May 2024. The digitization process leverages polarized directional lighting and neural networks to propagate microscopic details to a mesoscopic representation.

Use Cases

• Training neural networks for material appearance synthesis based on the described image-to-image translation methods.
• Validating BSDF models for realistic rendering based on the industry-standard models mentioned.
• Analyzing the optical properties of complex textiles based on features like yarn twist and fly-out fiber shape described in the data.
• Developing multi-scale digitization systems for materials based on the described device and methodology.

Strengths

• Includes data at two scales (micro and meso) for complex multi-scale materials.
• Captures anisotropic reflectance and transmittance using polarized directional lighting.
• Device is designed to digitize flat samples up to 11x11cm.
• Data is validated using 'challenging real-world material samples and validation scenes' as stated in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Garces, Elena via e-cienciaDatos Harvested Dataverse.

Collection Method

Acquired via a custom-built hemispherical light dome using polarized directional lighting and a single orthogonal camera.

Time Range

null

Freshness

Last updated 2024-05-05 07:31:27; freshness should be verified.

Geography

null