Skip to content

Maintainer and Site Operations

The public project identity is flavvesResearch. No personal author name is inferred or published by this documentation.

Repository metadata to set

These GitHub settings require repository administration access and are not stored in source files.

Description

Validate OpenCV checkerboard camera-calibration datasets for blur, coverage, pose diversity, duplicates, and reprojection error.

Topics

opencv
camera-calibration
computer-vision
checkerboard
dataset-quality
reprojection-error
camera-intrinsics
python

Homepage

https://flavvesresearch.github.io/opencv-calibration-audit/

Deploy GitHub Pages

The repository's Documentation workflow:

  1. builds MkDocs with strict warnings on pull requests;
  2. uploads a Pages artifact after a non-PR build;
  3. deploys only from the main branch;
  4. uses job-scoped minimum permissions.

In GitHub repository settings, choose GitHub Actions as the Pages source. After the first deployment, verify:

https://flavvesresearch.github.io/opencv-calibration-audit/
https://flavvesresearch.github.io/opencv-calibration-audit/sitemap.xml
https://flavvesresearch.github.io/opencv-calibration-audit/guide/validate-opencv-camera-calibration-dataset/

Configure Google Search Console

  1. Add https://flavvesresearch.github.io/opencv-calibration-audit/ as a URL-prefix property.
  2. Download the HTML verification file from Search Console.
  3. Place that exact file at the docs/ root, commit it, and let the documentation workflow deploy it.
  4. Submit https://flavvesresearch.github.io/opencv-calibration-audit/sitemap.xml.
  5. Use URL Inspection to request indexing for the home page and main dataset-validation guide.

Search indexing and structured data do not guarantee ranking.

Verify ChatGPT Search access

The published pages are static HTML with no login or client-side rendering requirement. After deployment:

  1. check the GitHub Pages domain-root robots.txt response;
  2. confirm it does not disallow OAI-SearchBot;
  3. confirm the hosting layer does not block OpenAI's published crawler IPs;
  4. request the home page and guide with a normal HTTP client and verify 200 responses.

For a GitHub Pages project site, a robots.txt under the repository subpath does not control the domain root. This repository therefore does not add a misleading project-level file.

OAI-SearchBot supports ChatGPT Search discovery. GPTBot concerns potential model training and can be managed independently. An llms.txt file is not a replacement for accessible content, normal links, canonical URLs, or a sitemap.

See OpenAI crawler documentation and ChatGPT Search guidance.

Complete the 0.2.2 release

After repository changes pass CI:

  1. publish version 0.2.2 through the existing GitHub Release and PyPI Trusted Publishing workflow;
  2. verify PyPI metadata exposes keywords and Documentation, Repository, Changelog, and Issues links;
  3. confirm the README's absolute report links work on PyPI;
  4. keep 0.2.0 and 0.2.1;
  5. yank 0.1.0.post9, 0.1.0.post10, and 0.1.2 with reason: Development scaffolding; use 0.2.0 or newer.

Yanking is a reversible PyPI-owner action and is not performed by the documentation workflow.

Technical sharing principles

Publish the main guide first and a real production benchmark only after a licensed dataset exists. When answering a relevant OpenCV Forum, Stack Overflow, or computer-vision community question, make the answer independently useful and link only where the audit directly supports the explanation. Do not create promotional questions, duplicate posts, or claims of guaranteed calibration quality.

Documentation version: 0.2.2 · Updated: 2026-07-31