thiagotigaz/ocr-it

What it solves

OCR It is a browser extension for Chrome and Firefox designed to extract text from paginated documents that are "trapped" in viewers—such as scanned books, slide decks, or PDFs—where text selection is disabled. It allows users to convert an entire multi-page document into a text file for use in LLMs or other applications.

How it works

Users define a capture region on the screen once. The extension then takes a screenshot of that specific area and performs Optical Character Recognition (OCR) using a bundled Tesseract build that runs entirely offline.

To handle multi-page documents, the extension can be configured to automatically turn the page using a stored screen point (to bypass cross-origin iframes and Shadow DOM) or a keyboard event. It can then repeat the capture-and-turn process in an automated "run" until the document ends, detecting the end of the document when text stops changing or a page cap is reached.

Who it’s for

It is for anyone needing to digitize text from web-based document viewers that prevent text copying, particularly those who want to maintain privacy by keeping all processing local to their machine.

Highlights

• 100% Offline OCR: Uses a bundled Tesseract build; no API keys or network requests are made.
• Automated Runs: Can automatically capture and turn pages in a sequence to digitize entire books.
• Cross-Frame Support: Uses coordinate-based clicking to interact with embedded viewers in cross-origin iframes or Shadow DOM.
• Local Export: Allows editing of captured text in-place and exporting the final result as a .txt file.
• Multi-language Support: Ships with English, Portuguese, and Spanish, with the ability to vendor in other Tesseract languages.