MakeOhio 3D Scanner — Smartphone Photogrammetry System
Cost-effective 3D scanning pipeline using smartphone photogrammetry, reducing scanning costs by 90% compared to professional-grade hardware.
90% cost reduction vs professional scanners ($50 smartphone vs $5,000+ hardware)
Sub-millimeter accuracy for objects under 50cm (validated against calibrated calipers)
Winner - Best Hardware Hack, MakeOhio 2025
Received NSF I-Corps grant for commercialization exploration
Professional 3D scanning equipment (LiDAR, structured light scanners) costs $5,000-$50,000, pricing out small manufacturers, educational institutions, and hobbyist makers from digital fabrication workflows. Existing photogrammetry tools require expensive DSLR cameras and complex calibration processes, making high-quality 3D capture inaccessible to most users.
We built a photogrammetry pipeline that converts smartphone camera captures into high-fidelity 3D models using computer vision algorithms. The system automates image alignment, depth reconstruction, and mesh generation—delivering professional-grade scans at 1/10th the cost of LiDAR scanners.
The Challenge
The democratization of digital fabrication (3D printing, CNC machining, laser cutting) has made manufacturing accessible to anyone with a few hundred dollars and internet access. But a critical bottleneck remains: capturing real-world objects as digital models. Professional 3D scanners deliver high accuracy but cost thousands to tens of thousands of dollars, limiting adoption to well-funded labs and corporate R&D departments.
Small manufacturers need to reverse-engineer replacement parts for legacy equipment. Educators want students to digitize sculptures for art history projects. Hobbyists want to replicate vintage car components for restoration. All face the same problem: scanning technology remains prohibitively expensive and technically complex.
Existing smartphone photogrammetry apps produce low-quality meshes riddled with holes and artifacts, suitable only for visual effects—not precision manufacturing. The gap between "free but unusable" and "accurate but $5,000" leaves a massive underserved market.
Our Approach
We developed a smartphone-native 3D scanning system optimized for accessibility without sacrificing quality:
-
Guided Capture Interface - React Native mobile app that uses augmented reality overlays to guide users through optimal camera positions. Computer vision algorithms detect insufficient coverage in real-time, prompting additional shots to eliminate scan gaps.
-
Cloud Processing Pipeline - Uploaded images are processed through a multi-stage photogrammetry pipeline: SIFT feature extraction for keypoint matching, structure-from-motion (SfM) for camera pose estimation, multi-view stereo (MVS) for dense point cloud generation, and Poisson surface reconstruction for watertight mesh output.
-
Automated Calibration - Self-calibrating camera intrinsic parameters from image metadata (focal length, sensor size) eliminates manual calibration steps. Automatic scale detection using fiducial markers or known object dimensions ensures dimensional accuracy.
-
Web Viewer - Three.js-based viewer allows users to inspect, measure, and export models (STL, OBJ, GLTF) directly from the browser. Integrated measurement tools display real-world dimensions and surface area calculations.
Results & Impact
Tested across 200+ scans ranging from mechanical parts to organic sculptures, the system achieved sub-millimeter accuracy for objects under 50cm—matching the performance of $5,000 structured light scanners while using only a $50 smartphone camera.
The project won Best Hardware Hack at MakeOhio 2025, competing against teams with dedicated engineering budgets. The judges highlighted the commercial viability of democratizing 3D scanning for educational and small-scale manufacturing applications.
Post-hackathon, the team received an NSF I-Corps grant to explore commercialization pathways, conducting 100+ customer discovery interviews with machine shops, museum conservators, and prosthetics manufacturers. Key insight: accuracy matters less than consistency—users prioritize reliable, repeatable scans over laboratory-grade precision.
The system is currently deployed in three Ohio high schools' STEM programs, where students use it to digitize historical artifacts for virtual museum exhibitions. One school reported a 300% increase in student engagement with 3D modeling assignments after introducing the accessible scanning workflow.
By reducing the cost barrier from thousands to effectively zero (most students already own capable smartphones), the project demonstrates how computer vision can unlock new markets by reframing expensive hardware problems as software optimization challenges.