Photogrammetry Workflow: From Physical Product to AR Experience

Why Photogrammetry Powers Modern AR Experiences

Photogrammetry transforms physical products into accurate 3D models by capturing them from multiple angles and computationally reconstructing their geometry and texture. For businesses deploying AR experiences, this approach delivers photorealistic assets at a fraction of traditional 3D modeling costs—particularly valuable when cataloging existing product lines or capturing unique items where CAD data doesn't exist.

The technology excels for products with consistent, non-reflective surfaces: furniture, packaged goods, sculptures, machinery, and architectural elements. While certain materials present challenges, understanding the workflow enables businesses to determine whether photogrammetry suits their product catalogue and budget constraints.

Equipment Requirements and Investment Levels

Photogrammetry setups scale from DIY configurations under £2,000 to professional rigs exceeding £10,000. For most SME applications, mid-range equipment delivers excellent results:

• Camera: DSLR or mirrorless with 20+ megapixels (£500-1,500) provides sufficient detail for products up to 2 meters
• Lighting: LED panels or softboxes (£200-800) ensuring even, diffused illumination without harsh shadows or hotspots
• Turntable: Motorized rotation platform (£150-600) for consistent angular spacing between captures
• Processing software: Solutions like RealityCapture or Metashape (£150-3,500 annual licenses) handle reconstruction
• Optimization tools: Blender (free) or specialized software for polygon reduction and texture baking

Alternatively, outsourcing to professional photogrammetry services costs £200-1,000 per model depending on complexity, eliminating equipment investment while providing guaranteed quality for businesses requiring occasional 3D assets rather than continuous production.

Capture Techniques for Different Materials

Material properties dramatically impact photogrammetry success. Ideal subjects have: matte or slightly textured surfaces, consistent coloration without transparency, and distinctive surface features enabling software to identify matching points across images. Challenging materials require specialized techniques or alternative approaches.

Common material challenges and solutions include:

• Reflective surfaces: Apply temporary matte dulling spray that washes off, or cross-polarized lighting to minimize specular reflections
• Transparent materials: Often require manual 3D modeling or combination approaches—photogrammetry for opaque components only
• Pure white or black objects: Add temporary surface markers or patterns to provide tracking points for reconstruction software
• Large products: Break into manageable sections captured separately, then digitally assembled in optimization phase

Processing Workflow and Optimization Pipeline

Raw photogrammetry output requires significant optimization before AR deployment. Initial reconstructions often contain millions of polygons and multiple 4K textures—far exceeding mobile device capabilities. The optimization workflow transforms these dense meshes into efficient AR assets:

• Polygon reduction: Target 20,000-100,000 triangles depending on product complexity and platform requirements
• Texture baking: Consolidate high-resolution surface detail into single 2K textures through normal and ambient occlusion mapping
• UV optimization: Ensure texture coordinates maximize texture space utilization while minimizing stretching
• LOD creation: Generate multiple detail levels automatically switching based on viewing distance
• File format conversion: Export to platform-specific formats (USDZ for iOS, glTF for Android and WebAR)

Quality Benchmarks and Platform Specifications

AR platforms impose specific technical requirements impacting optimization targets. iOS ARKit and Android ARCore recommend: maximum 100,000 triangles per model, single 2K texture maps, PBR material definitions using metallic-roughness workflow, and total file sizes under 10MB for acceptable loading times on mobile connections.

Quality assessment focuses on geometric accuracy (within 1-2% of physical dimensions), texture fidelity (sufficient resolution to represent product details at typical viewing distances), and absence of reconstruction artifacts (holes, floating geometry, or incorrect surface connections requiring manual cleanup).

When to Use Photogrammetry Versus Manual Modeling

Photogrammetry excels for products with organic shapes, complex surface details, or unique items where capturing existing objects proves faster than recreating from scratch. Consider manual 3D modeling when: products don't physically exist yet (prototypes or concepts), materials prove impossible to photograph effectively (highly reflective jewelry, transparent glassware), or products require frequent variations where parametric modeling enables rapid option generation.

Many businesses adopt hybrid workflows—photogrammetry for base assets combined with manual modeling for problematic components or customizable elements. This pragmatic approach optimizes both cost and quality while accommodating diverse product characteristics within a single AR catalogue.