How to digitise a fabric library: scanning, tagging and AI search

What hardware digitises fabric libraries?

Dedicated fabric scanners combine high resolution cameras with controlled lighting to capture texture, colour and weave structure. Modern fabric scanners typically combine high resolution cameras with uniform lighting systems to eliminate shadows and glare, ensuring accurate colour representation, and the captured data generates comprehensive texture maps that allow computers to visualise materials in 3D with remarkable fidelity to their physical counterparts.

Devices like Vizoo and X-Rite's Tac7 dominated the textile digitisation market for nearly a decade, producing 3D ready material data that powered fashion design, apparel prototyping, and product visualisation. Vizoo, founded in 2013 in Germany, focuses on high resolution 3D scanning and texture creation, and their flagship xTex Scanner captures exceptionally detailed digital representations of fabrics using Nikon's Z7 camera series with xTex Software as a comprehensive material editor for tiling, recolouring, and exporting texture maps. Other established vendors include NunoX, Colormass and DMix.

The future of surface digitisation no longer depends on scanning hardware, being rewritten by generative AI systems that can infer depth, normal, and specular data from a single 2D fabric image. Traditional scanners required calibrated cameras, controlled lighting, and expensive workflows, and while accurate, they were limited by hardware costs, environmental sensitivity, and a dependence on physical samples. AI driven approaches now allow fashion teams to capture fabric swatches with flatbed scanners or smartphones and generate the missing 3D material parameters through inference.

Which file formats support interoperability?

Fashion's ambitions are held back by a persistent lack of clarity and alignment around who should handle the work of scanning fabrics, what targets and quality profiles they should work to, where the ongoing cost of digitisation should reside, and what standards can provide the best coverage for allowing different 3D tools, product lifecycle management platforms, and other systems to extract value from and add value to those files.

MaterialX is an open source standard for describing, authoring, and exchanging rich material and shader networks across 3D content creation applications and rendering platforms, originally developed by Lucasfilm and Industrial Light & Magic, providing a standardised framework for representing shading graphs, procedural textures, and material definitions in a renderer independent format. OpenPBR Surface is the default surface shader as of Maya 2026, with MaterialX 1.39.4 integrated in Maya 2027, and OpenPBR Surface is the default material as of 3ds Max 2026.

OpenPBR is a joint effort between Adobe and Autodesk, developed under the guidance of the MaterialX Technical Steering Committee and released through the Academy Software Foundation, built specifically as the successor to Adobe Standard Material and Autodesk Standard Surface, essentially merging two competing internal standards into one open, vendor neutral specification. Substance 3D Painter, Maya 2025.3, and 3ds Max 2026 all use OpenPBR as their default material, working with it just means working with the standard shader.

How do you tag materials for search?

Metadata determines whether a digitised fabric can be found. A searchable material library requires structured data fields that describe composition, physical properties, supplier, certifications and visual attributes. The capturing process itself is reasonably well defined, but the infrastructure required to undertake it continuously and consistently exists in a patchwork state, between individual mills and material suppliers, regionalised hubs hosted and sponsored by digital materials libraries, and brands' own scanning efforts, and if the assets need to have utility beyond visual representation and to carry embedded datapoints like test and certification results, finishing details, yarn and fibre specifications, construction details, shrinkage, production provenance and more, then those assets need to not only become richer, but to become standardised in a way that ensure they can flow through the entire product journey, from 3D CAD, through PLM and enterprise resource planning.

Effective tagging schemas include fibre composition, weight in grammes per square metre, weave or knit structure, colour values, finish, stretch percentage, care requirements, certifications such as GOTS or OEKO TEX, supplier name, and minimum order quantity. Digital materials are mainly created to serve the most mature use cases, which all centre on visualisation, and as a result of this emphasis on how a fabric looks, visually, data is either left on the table at the capture stage or isn't factored.

What role does AI play in material search?

AI fabric finders are changing how designers, manufacturers, and retailers identify textiles by analysing fabric images with computer vision and machine learning, and instead of relying solely on manual inspection, these tools recognise textures, weave structures, colours, and material characteristics to deliver fast and consistent identification results, supporting textile sourcing, quality control, digital fashion, and sustainable product development by improving accuracy while reducing manual effort.

AI fabric finders represent a breakthrough in textile identification technology, using machine learning algorithms to scan uploaded photos and match them against vast databases of fabric types, whether dealing with cotton weaves, synthetic blends, or luxury silks, these systems process intricate details such as weave patterns, fibre composition, and surface sheen in seconds, eliminating guesswork and making fabric recognition faster and more reliable than traditional manual methods. Semantic search allows designers to query a library with natural language such as 'lightweight linen for summer shirting' or find visually similar materials by uploading a reference image.

Who provides material digitisation services?

Mills, brands and specialist service providers all digitise fabric libraries. From its launch in 2020, the Cotton Incorporated digital materials library has been an outlier, a free resource, filled with interoperable materials, in a landscape defined by closed platforms and proprietary standards, and today, the library encompasses more than 500 high fidelity fabrics in both native CLO and Browzwear formats that accurately represent the composition and characteristics of real cotton or cotton rich materials.

PVH Corp, which owns Tommy Hilfiger and Calvin Klein, has heavily invested in 3D design and digital sampling, aiming to eliminate physical samples from the design process, and by integrating digital twins, PVH has cut sample volumes by 60 per cent. Centralised digitisation reduces duplication and ensures consistency, but the cost and responsibility for maintaining a current, accurate library remain points of tension between brands and their mill partners.

Digital transformation is no longer a future goal, but a baseline for survival in 2026, driven by global trade shifts and consumer demand for speed and customisation, and the modern printed textile supply chain relies on a seamless digital thread that connects every stage of production, ensuring data flows uninterrupted from the designer's screen or buyer's desk to the shipping dock.

What are the practical steps to digitise a library?

Begin by auditing the physical collection. Prioritise fabrics that appear frequently in designs or represent key supplier relationships. For each material, capture a high resolution scan or photograph under daylight equivalent illumination. Record structured metadata in a consistent schema, including all physical test results and supplier information.

Choose whether to invest in dedicated scanning hardware or use AI inference tools that generate 3D material properties from 2D images. The price of the hardware required to capture materials, and the skills required to use it, also scale with the intended use case: a simple visual scan can be accomplished on consumer grade hardware, or even smartphones, but archival quality scans require expensive, complex machinery to capture.

Export files in open formats such as MaterialX or vendor native formats for CLO, Browzwear or other 3D design tools. Store digital assets in a cloud based digital asset management system with robust search and version control. Cotton Incorporated sees the most value in pairing its library with renewed investment in support for material standards, educating suppliers, and taking part in bilateral industry dialogues. Integrate the library with PLM and design tools to ensure designers can access materials directly within their workflow rather than switching between systems.

Researched and drafted with AI support, reviewed and released by the editorial team.

EN DE FR IT