Image checks at capture
A photo checked for blur or framing before it joins the record.
We have no published computer vision build. Our inspection apps capture photos and signatures as evidence, and we show where OpenCV would check an image before it reaches the record.
A photo checked for blur or framing before it joins the record.
Photos kept with GPS, time and readings, as on the Digital Testing Platform.
Thresholds and exception workflows around any automated check.
Moves between 4.x and 5.0.
Illustration. A dashed outline marks where the technology would sit. We have no published build on it.
Source: github.com/opencv/opencv, checked 4 November 2026. The releases page shows day and month without a year.

Photo and signature evidence from the field, with validation rules and certificates on completion.
A photo and a signature required before a drop can close, which cut paperwork disputes.
Python
YOLO
PyTorch
TensorFlow
Docker
iOSThe repository shows the Apache-2.0 licence.
No. We say so on the first call and show the nearest work, which is photo and signature evidence capture on inspection apps.
GitHub marks OpenCV 5.0.0 as the latest release, and 4.14.0 is the newest release on the 4.x line.
You do, from the first commit. Repositories and cloud accounts stay in your name.
Thirty minutes on your images, your rules and your inspection process. No sales pitch.