Card Front Grader

Batch process images, compute centering + heuristic flaw metrics, and render overlays.
Next.js + OpenCV.js

1) Upload images (front only)

Upload a card photo, review the detected measurement area, then estimate the grade once the guides look right.
Nothing is persisted automatically. Images and overlays stay in memory for the current session unless you explicitly export results.
Drag and drop images here
Filters to PNG, JPG, and WEBP. Uploading now prepares the card bounds for review instead of immediately grading the image.
Status flow: PENDING - PREPARING - REVIEW - PROCESSING - DONE/ERROR.

2) Queue

Select a row to review the overlay, adjust the measurement area, and then estimate the grade for that image.
FileStatusEngineAIElapsedGradeEst. #CenterFlaw ptsConf.
No files yet.

3) Review and estimate

Select an uploaded image to review its measurement guides.