Fix low-quality OCR before approval

Some pages are not worth auto-processing twice—know when a new photo beats another automated pass.

Low-quality OCR is fixed or rejected before approval. You stop patching hopeless captures and know when a new photo beats another automated pass.

How to do it

Work through one batch at a time. Finish review before you upload the next set of pages.

Flag captures with blur, shadow, or cut-off characters for re-shooting first.
Re-run extraction only after improving capture quality—not repeatedly on the same bad photo.
Compare each draft field to the source; edit only when the underlying text is unambiguous.
Reject cards where OCR invented characters you cannot verify on the image.
For mixed layouts, split Thai and English into predictable front/back roles.
Approve the reduced, trustworthy set; export or study without carrying OCR debt forward.

Same error after re-run

the photo is still bad—lighting or crop must change.

Partial line readable

crop tighter to the readable region instead of approving the whole block.

Too many rejects

slow down capture habits before scaling uploads again.

Continue with nearby topics in the same workflow—import, review, then study or export.

Pillar

Convert Thai textbook photos into reviewed Anki cards

Next in lane

Continue this lane: Build a weekly vocab deck from class notes

Cross-topic A

Compare approaches: One-click generators vs editable review systems

Cross-topic B

Key concept: Cloze vs basic card formats

Start with the pillar that matches your question, then follow the linked guides for the next step.

Thai flashcards

How messy Thai source material becomes reviewed draft cards you can trust.

Convert textbook to Anki cards

The capture and extraction path from photos to reviewed cards.

Anki export

Move approved cards cleanly into an APKG-based Anki workflow.

Upload your study pages, review the extracted cards, and export only the material you want to keep in Anki.