PgsToSrt

Convert pgs subtitles to srt using OCR.

Prerequisites

• .NET 5.0 Runtime
• Tesseract 4 language data files

Data files must be placed in a tessdata folder inside PgsToSrt folder, or the path can be specified in the command line with the --tesseractdata parameter.

You only need data files for the language(s) you want to convert.

Usage

dotnet PgsToSrt.dll [parameters]

Parameters
--input	Input filename, can be an mkv file or pgs subtitle extracted to a .sup file with mkvextract.
--output	Output Subrip (.srt) filename. Auto generated from input filename if not set.
--track	Track number of the subtitle to process in an mkv file (only required when input is a matroska file) This can be obtained with mkvinfo
--trackLanguage	Convert all tracks of the specified language (only works with mkv input)
--tesseractlanguage	Tesseract language to use if multiple languages are available in the tesseract data directory.
--tesseractdata	Path of tesseract language data files, by default 'tessdata' in the executable directory.

Example (Command Line)

dotnet PgsToSrt.dll --input video1.fr.sup --output video1.fr.srt --tesseractlanguage fra
dotnet PgsToSrt.dll --input video1.mkv --output video1.srt --track 4

Example (Docker)

docker run -it --rm -v /data:/data \
           -e INPUT=/data/myImageSubtitle.sup \
           -e OUTPUT=/data/myTextSubtitle.srt \
           -e LANGUAGE=eng \
           tentacule/pgstosrt

Dependencies

• Windows : none, tesseract/leptonica libraries are included in the release package.
• Linux : libtesseract4 (sudo apt install libtesseract4)

Build

To build PgsToSrt.dll execute this commands

dotnet restore
dotnet publish -c Release -o out
#The file is on  PgsToSrt/out/PgsToSrt.dll

To build docker image (for now only for linux)

docker build -t pgstosrt .

Built With

• LibSE from Subtitle Edit
• Tesseract .net wrapper
• CommandLineParser
• SixLabors ImageSharp