Research figures & supplements: make captions searchable
Route `tiff-ocr-research` (tiff_to_text.research) targets paper figures, supplements, and publisher-grade TIFFs. Extracting methods blurbs or legend abbreviations must respect copyright: third-party figure text still follows the publisher’s reuse rules. Cross-check OCR against searchable PDFs when available; treat equations and Greek symbols as typography, not OCR-only tasks.
TIFF OCR workflow for research assets
- Upload allowed figure TIFFs and crop legend or methods blocks before recognition.
- Compare transcripts to the PDF or author manuscript to fix journal-specific abbreviations and superscripts.
- Note DOI, panel IDs, and license terms in your lab notebook for future citations.
TIFF-to-text FAQ (research)
What compliance issues appear when using `tiff-ocr-research` on someone else’s figures?
Honor copyright and translation policies; fair-use boundaries apply to teaching or commentary, republication needs permission and attribution.
When OCR disagrees with the publisher PDF, which source wins?
Treat the searchable PDF or author proof as canonical; use OCR from TIFF only after principal-investigator review when raster is all you have.
How do co-authors avoid overwriting the same figure caption?
Version filenames with DOI, panel ID, and date; lock milestone folders read-only.
May I paste OCR’d methods text straight into a thesis?
Only if it is your original prose or licensed; quote third parties with citations and figure IDs.
Greek letters and units break across lines—what is the pragmatic fix?
Transcribe symbol-dense lines manually or in LaTeX; reserve OCR for plain sentences.