OCR extraction

Extract Burned-in Subtitles from Video

Turn subtitles burned into video frames into editable SRT or VTT files with OCR and accurate timing.

AI OCR for burned-in captions
Read subtitles that are permanently rendered into video frames and turn them into editable, timed subtitle files.
  • Detect captions in the picture even when no subtitle track exists
  • Recover cue timing and export SRT, VTT, or ASS
  • Review OCR text, line breaks, and timing before export

Would you use OCR to recover subtitles from video?

Tell us which video formats, languages, or subtitle layouts matter most.

Was this tool helpful?

A quick thumbs up or down helps us improve this tool.

Buy me a Coffee

Extract hardcoded (burned-in) subtitles from video to SRT or VTT

Extract hardcoded subtitles from video with OCR. Read text permanently rendered into the picture, recover subtitle timing, and turn burned-in captions into editable SRT, VTT, or ASS files for translation, editing, and playback.

Example input and output

Input video
MP4, MOV, WebM, or MKV Burned-in Chinese or English subtitles
Editable output
SRT / VTT / ASS Recognized text with recovered cue timing

Best for

Recover a missing subtitle track

Create editable subtitles when the original SRT, VTT, or ASS file is unavailable and the only source is a video with visible captions.

Translate burned-in captions

Extract the source-language text, review OCR results, and send the recovered subtitle file into a translation workflow.

Make on-screen text searchable

Turn captions from lectures, interviews, clips, and archived video into searchable text for review and cataloging.

Common file issues handled

Review text before export

Correct OCR mistakes, split or merge lines, and check subtitle timing before downloading the recovered file.

OCR works on visible captions

Recognition depends on resolution, contrast, font style, animation, and whether captions are clear against the background.

Precise timecodes

Frame sampling and subtitle changes are used to build timed cues that stay aligned with the source video.

Common subtitle formats

Export SRT for players and editors, VTT for web video, or ASS for styled desktop playback.

Related workflows

Frequently asked questions

What are burned-in subtitles?
They are subtitles permanently rendered into the video image. Unlike a separate SRT or VTT track, they cannot be switched off in a player.
What can the extractor export?
The OCR workflow is designed to recognize subtitle text from video frames, recover cue timing, and export editable SRT, VTT, or ASS subtitles for review.
Can I extract subtitles from an MP4 or MOV file?
Yes. The workflow is designed for common video containers such as MP4, MOV, WebM, and MKV when the subtitle text is visible in the video image.
What is the difference between hardcoded and embedded subtitles?
Hardcoded or burned-in subtitles are pixels rendered into each video frame. Embedded subtitle tracks are separate streams that can usually be extracted without OCR.
Can I edit the extracted subtitle text?
Yes. OCR results should be reviewed for names, punctuation, line breaks, and timing before you download or translate the subtitle file.