Scan4D

Capture, re-scan, and export high-fidelity 3D scans with LiDAR depth, RGB, and camera poses — all from your iPhone or iPad.

Scan4D is a professional-grade reality capture app for iPhone and iPad. It turns your device's LiDAR sensor and camera into a powerful 3D scanner, capturing detailed meshes, depth maps, and camera poses for downstream reconstruction, photogrammetry, and spatial computing.

FULL MODE (LiDAR devices) • Real-time mesh reconstruction with live wireframe overlay • 16-bit depth maps captured at adaptive frame rates • Scan capacity metrics: polygon count, drift tracking, session duration • Interactive 3D mesh preview with camera-sampled vertex coloring • Privacy filtering: person segmentation removes humans from meshes, face blurring in RGB exports

LITE MODE (all devices) • RGB frame capture with ARKit camera poses • Server-side photogrammetry via exported image bundles • Automatic detection — no configuration needed

TIME-SERIES SCANNING • Group scans by Location and re-scan the same space over time • "Extend Scan" loads a red ghost-mesh overlay of previous captures • Adjacent stitching: scan neighboring areas with shared coordinate frames • ARWorldMap relocalization for precise alignment between sessions

FLEXIBLE EXPORT FORMATS • Scan4D: Full bundle with metadata, relocalization map, images, depth, and camera data • Polycam: Raw data import compatible with Polycam • RAW: Nerfstudio/COLMAP-compatible transforms.json + images + depth • OBJ / PLY / USDZ: Native mesh formats for MeshLab, Blender, and iOS Quick Look

SERVER INTEGRATION • Configure any HTTP(S) upload endpoint in Settings • Direct PUT upload with live progress tracking • Save to Files or AirDrop for offline workflows • Compatible with self-hosted reconstruction pipelines

DEVELOPER MODE • Front/back camera switching for testing privacy features • Synthetic IMU, camera, and depth injection for Simulator testing • Vertex mapping diagnostics and debug overlays

Designed for researchers, 3D scanning professionals, and spatial computing developers who need raw, high-fidelity sensor data with full control over the reconstruction pipeline.

Requires ARKit-capable device. LiDAR-equipped iPhone or iPad Pro recommended for full functionality.