VellumVision

Restore and reconstruct damaged medieval manuscript folios with AI

Image GenerationPaid

VellumVision applies inpainting models trained on undamaged folio scans to reconstruct lacunae, water damage, and foxing obscuring illuminated text. The system cross-references known scribal hands and regional iconographic conventions to propose textually and visually coherent infill. Conservation teams receive confidence-scored restoration previews alongside detailed provenance-safe audit trails.

Key Features

• ✓AI inpainting for lacunae and water damage
• ✓Scribal hand recognition across 400+ identified hands
• ✓Iconographic convention cross-referencing by region
• ✓Confidence-scored restoration preview rendering
• ✓Provenance-safe audit trail generation