OCR automation for VideoSubFinder.
pip install git+https://github.com/moi15moi/VideoSubOCR.git
- Python 3.7 or more
- ABBYY FineReader (or Tesseract)
- Tesseract (or ABBYY FineReader)
- VapourSynth (Optional, but required to use cropbox argument)
- VapourSynth-Editor (Optional, but required to use VideoPreview.vpy)
Setup for ABBYY FineReader
You will need to have the same parameters has this image.
To change your parameters, open ABBYY FineReader and go to Tools -> Options -> Format Settings -> TXT
Setup for Tesseract
To be able to select the language you want, you need to download tesseract tessdata.
I recommand you to use tessdata_best traineddata.
You will need to download the language you need and move it in the right folder.
On Windows, it should be: C:\Users\YOUR_USERNAME\AppData\Local\Tesseract-OCR\tessdata
usage: VideoSubOCR [-h] --input INPUT [--generalSettings GENERALSETTINGS]
[--cropBoxDimension Width_or_Height Width_or_Height] [--cropBoxHeight Height]
[--videosubfinderwxw VIDEOSUBFINDERWXW] [--tesseract [TESSERACT]] [--finereader [FINEREADER]]
[--lang LANG]
Video OCR automation.
options:
-h, --help show this help message and exit
--input INPUT, -i INPUT
Video file needed to be OCR.
--generalSettings GENERALSETTINGS
Path to general settings (*.cfg file). By default, VideoSubFinder use the file
settings/general.cfg
--cropBoxDimension Width_or_Height Width_or_Height, -cpdh Width_or_Height Width_or_Height
CropBox dimension Width x Height. Ex: --cropBoxDimension 1920 1080
--cropBoxHeight Height, -cph Height
CropBox height. It is the height between the video bottom and the cropbox bottom.
--videosubfinderwxw VIDEOSUBFINDERWXW, -vsf VIDEOSUBFINDERWXW
Path to VideoSubFinderWXW.exe if not in variable environments.
--tesseract [TESSERACT], -t [TESSERACT]
If tesseract.exe is in your variable environments, simply use --tesseract. If not, path to
FineCmd.exe
--finereader [FINEREADER], -f [FINEREADER]
If FineCmd.exe is in your variable environments, simply use --finereader. If not, path to
FineCmd.exe
--lang LANG, -l LANG Language in which the hardsubbed subtitle is. It will be use by the ocr engine you choose.
Warning, ABBYY FineReader and Tesseract doesn't have the exact same input for the same
language. Tesseract support ISO 639-2 (t version):
https://www.loc.gov/standards/iso639-2/php/code_list.php ABBYY FineReader support the
fullname: https://help.abbyy.com/en-us/finereader/15/user_guide/commandline_lang/
I recommand you to always create a setting.
This will allow to have a better ocr
You can save the settings with File --> Save Settings As...
For more information about VideoSubFinder, see their website
If you don't want to create a setting, I recommand you to use the file named VideoPreview.vpy. To use it, you will need to install VapourSynth and VapourSynth-Editor.
- Open VideoPreview.vpy with VapourSynth-Editor.
- Change
inputVideo
,cropBoxDimension
,cropBoxHeight
to the value you think you need. - Press F5
- See if the cropbox is good for you. If not, repeat the step 2 and 3.