With the development of 3D rendering technology and the industry’s growing demand for high-quality digital modeling, more and more 3D applications are adopting physically based rendering (PBR). Realistic rendering requires support from high-quality texture maps, primarily including base color maps, normal maps, metallic maps, roughness maps, and transparency maps.
Most existing material-scanning technologies rely on image-processing methods, such as overlaying, subtracting, inverting, blurring, and filtering images captured under different lighting conditions, in order to obtain texture maps.
Although this approach can produce certain effects, it still has several obvious problems: it is easily affected by the fabric’s own color, leading to deviations in texture analysis; it does not analyze specular information sufficiently, especially in the representation of metallic and roughness properties; and the generated map types are often incomplete, making them difficult to adapt to high-quality PBR rendering systems.
The final result is that a digital fabric may have texture, but lacks true materiality.
The patent introduced in this issue addresses precisely the problem of restoring realism in digital fabrics.
Unlike traditional scanning technologies that rely on post-production processing of images, Method and electronic device for generating texture maps through multi-angle lighting capture (US12368968B2) and Method for generating material map by means of multi-angle lighting photography (EP4290460) are based entirely on physically based rendering principles. The patents can generate high-quality base color, normal, metallic, roughness, and transparency maps in a single workflow.
Rather than simply “scanning an image,” the technology restores the full physical material information of a fabric. One particularly important breakthrough is its handling of intrinsic color interference.
In conventional scanning processes, the fabric’s own color often affects the analysis of highlights and roughness. For example, dark and light fabrics produce different visual feedback under illumination, which can interfere with the system’s judgment of the true material parameters.
This technology subtracts the base color from the color values of multi-lighting images, effectively removing the influence of intrinsic color and significantly improving the accuracy of roughness and metallic parameters. This means the system can more accurately identify the fabric’s true reflectance and surface structure, rather than being misled by the color itself.
The patent also designs a calibration workflow using a standard color chart and white paper to calibrate the color matrix and light-intensity coefficients. Because the calibration data can be reused, the workflow can effectively compensate for the effects of light-source distance and angle. As a result, fabric scanning becomes less dependent on professional manual color tuning and experience-based judgment, shifting toward a more standardized and stable digital process.
This higher-precision and more standardized digital-fabric capability determines its application value in real commercial scenarios.
In fields such as virtual try-on, 3D garment design, games, and e-commerce presentation, fabric materiality is often one of the most critical factors affecting the final visual effect. The texture maps generated by this patent are compatible with most mainstream rendering systems today, enabling enterprises to migrate physical fabric assets into digital form at lower cost. This capability carries significant value for equipment sales, SaaS services, and technology licensing.
More broadly, digital fabric technology forms an important part of Style3D’s digital product creation stack, working alongside garment simulation, rendering, and AI-assisted generation to support more realistic and production-ready digital garments.
To learn more about how AI and 3D technology can streamline your workflow and bring digital fabrics to life, please contact us.
