An Optimized Workflow for Material Texture Reconstruction

Prototyping is a crucial phase during product development that allows designers to evaluate and refine concepts through iterative processes. Projection-based spatial augmented reality (P-SAR) has significantly enhanced this phase by enabling mixed prototyping, which combines tangible physical prototypes with virtual enhancements. This approach facilitates real-time manipulation of perceptual features, such as surface material appearance and shape deformation, without requiring physical displays or additional electronics. While P-SAR technology has the potential to accelerate design iterations, reduce costs, and improve collaborative communication, it faces critical barriers to widespread adoption. These include complex setups, the expertise required for content creation, and the challenges of realistically reproducing material surfaces. Additionally, the reliance on technical experts and professional-grade equipment for scanning and digitising physical materials increases the time and cost of content creation, diminishing the advantages of mixed prototyping. Addressing these issues requires more intuitive and efficient material scanning tools and workflows that ensure accurate representation while reducing the complexity of the digitisation process. To address these challenges, this work proposes an optimised workflow for scanning materials and generating their PBR (Physics-Based Rendering) textures, enhancing the realism of mixed prototypes in subsequent P-SAR applications. Specifically, the workflow targets a variety of materials commonly used during the product prototyping phase, such as fabrics and plastics. It begins with a custom-designed scanner that can be configured to the specific properties of the material being scanned. The workflow integrates photometric stereo-based raw image data acquisition with an adaptable software post-processing pipeline that outputs albedo and normal maps, with optional extensions for additional maps as required.