Skip to content

projectSetup.md

Latest commit

 

History

History
68 lines (53 loc) · 2.6 KB

projectSetup.md

File metadata and controls

68 lines (53 loc) · 2.6 KB
Raw

🔍 Veritas — Development & Setup Notes

Short setup guide. For the feature overview see README.md; for module APIs see docs/API.md.

1. Environment

Python 3.10 (pinned in .python-version).

git clone https://github.com/CodeRafay/Forensic-Image-Analysis-Toolkit.git
cd Forensic-Image-Analysis-Toolkit

python -m venv .venv
# Windows
.venv\Scripts\activate
# macOS/Linux
source .venv/bin/activate

pip install -r requirements.txt

requirements.txt pins the versions the tests were run against: streamlit, numpy, scipy, scikit-image, opencv-python-headless, Pillow, PyWavelets, matplotlib, piexif, ImageHash, c2pa-python. requirements-dev.txt holds optional linters and is not needed to run the app or the tests.

2. Layout

app.py               Streamlit UI: sidebar upload, 13 tabs, one render() for all results
analysis/            One module per technique; util.py holds the result contract
Descriptions/        Per-technique guides shown in each tab
tests/               test_<module>.py per module + test_contract.py (142 tests)
assets/              style.css, sample images/sampleImg.jpeg (fabricated demo image)
.streamlit/          config.toml (dark theme, headless, showErrorDetails="type")
docs/                API, deployment, project report

3. How app.py works

  • Uploads are written to a per-session temporary directory (tempfile.mkdtemp(prefix="veritas_")) under a random file name. The previous upload is deleted when a new one arrives; results are cleared when the image changes (tracked by SHA-256 of the upload).
  • No preprocessing. Every tab receives the path of the original file. Nothing is downscaled or re-saved first. cmfd.detect_copy_move downscales in memory to 2048 px (HEAVY_MAX_PX); prnu.analyze_prnu centre-crops.
  • One renderer. run_panel(key, label, fn, *args, **kwargs) draws a button, calls the analysis, caches the result in st.session_state.results and passes it to render(result). A module exception becomes an error result instead of crashing the page.
  • Disclaimer. An "indicators, not proof" notice sits under the title; the Synthetic-traces tab carries an additional Experimental warning.
  • Hash ledger lives in st.session_state.ledger; exports are signed with HMAC-SHA256 when st.secrets["LEDGER_KEY"] exists.

4. Run

streamlit run app.py                                  # http://localhost:8501
python -m unittest discover -s tests -t .             # full suite, a few minutes

5. Deploy

See docs/DEPLOYMENT.md (Streamlit Community Cloud, Docker).