This userscript is designed to help you extract hardcoded subtitles from YouTube videos using Optical Character Recognition (OCR). The script provides an option to crop a specific area of the video for OCR processing.
- Crop Area Selection: Select a specific area of the video to focus the OCR process.
- Automatic Pausing: The video automatically pauses during the OCR process to ensure accurate text recognition.
Userscript Manager for browser, visit greasyfork for installation instructions.
-
Open browser and enable extension
-
Download and install YouTube Sub OCR UserScript
- Download from GitHub yt-sub-ocr.user.js
- Open a YouTube video.
- Click on the
Crop
button below video canvas to select the area for OCR. - Click the
OCR
button to begin the OCR process. - The video will pause, and the selected area will be processed for text extraction.
- The recognized text will be displayed below the video canvas.
- The OCR is done by paddlejs-ocr
- The crop area is done by cropperjs