Computer Vision · 2026
Walet Length Calculation
Calibrated image measurement with 0.01–0.05 cm reported error
A calibrated vision system that turns images into physical measurements with ChArUco, ArUco markers, and automated scale conversion.
- Python
- OpenCV
- ChArUco
- ArUco
- Vercel
Measured evidence
- reported measurement error
- 0.01–0.05 cm
- Improved from approximately 0.5–1 cm.
Problem worth solving
Estimate physical length from images while controlling distortion and scale errors that made the initial measurements unreliable.
Engineering decisions
- Applied checkerboard and ChArUco camera calibration to reduce lens-distortion error.
- Automated ArUco detection and pixel-to-length conversion.
- Packaged the workflow as a web application deployed on Vercel.
What changed
- Reduced the reported measurement error from roughly 0.5–1 cm to around 0.01–0.05 cm.
- Turned the calibrated workflow into a deployed image-measurement application.