Pokemon Border Grader (MVP)

This is a first-pass grader for Pokemon card centering.

Current scope:

• Detects the card in an uploaded image.
• Estimates border thickness on all four sides.
• Reports left/right and top/bottom border ratios.
• Flags pass/fail for a 45/55 centering window.

Setup

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Run

streamlit run app.py

Then upload a card image in the browser UI.

Notes

• This is a heuristic model, not an official TAG-grade implementation.
• Image quality, glare, and angled photos can affect results.
• Best results come from flat, well-lit scans/photos with clear card edges.