Adapted (aka shamelessly stolen) from the bash based video OCR YolOCR
https://git.clapity.eu/Id/YoloCR or https://bitbucket.org/YuriZero/yolocr/src/master/
It uses vapoursynth, ffmpeg and tesseract along with some python modules Note: A fork of https://github.com/Lypheo/pythOCR originally from https://github.com/pocketsnizort/pythOCR
- It doesn’t depend on pyEnchant anymore, which has been discontinued and doesn’t have any 64 Windows builds.
- It allows the filter parameters in the vpy file to be set via command line arguments of the main script.
- It’s likely even more broken than the original because I can’t code for shit.
11/03/2022:
- Cleaned code, updated deprecated functions, updated libs to 64 bit
TODO:
- Cross check the code with the original repo
- Fix path error
- Migrate to click or cloup from configargparse
- Make colored text
- Add more detailed docs
- ...
Install
- python 3+ (Use the 32-bit version on Windows, because pyEnchant does not support Win64)
- vapoursynth
- tesseract. (Make sure to use the official training data set. The one included in the Windows setup by UB Mannheim is faulty and needs to be replaced.)
Install vapoursynth plugins:
- For Windows:
- Copy the content of dependecies/Win/vapoursynth_dep to your vapoursynth plugin folder
- Copy the content of dependecies/Win/python_dep to your python site-packages folder
- For Linux:
- install Vapoursynth-plugins (or use my containerized version of pythOCR)
Install (from pip) python prerequisites:
- Linux
$pip3 install colorama configargparse pyspellchecker numpy opencv-python tqdm typing - Windows
pip install colorama configargparse pyspellchecker numpy opencv-python tqdm typing
clone this repository
Here, you're done.
$python3 pythOCR.py --help
usage: pythOCR.py [-h] [-V] [-c CONFIG] [-l language] [-wd folder] [-o folder]
[-log level] [-ass style] [-rr path to regex-replace json]
[-hcr char,replace] [-sf format] [-m mode] [-vpy vpy_file]
[-T number] [-autosamesub number] [-samesub number] [-nsc]
[-t] [-d] [-tss path to tesseract binary]
[-vps path to vspipe binary] [-wdt number] [-hgt number]
[-cropy number] [-cropalty number] [-ss number] [-er number]
[-rs mode] [-wt number] [-bt number] [-dt number]
path [path ...]
Filters a video and extracts subtitles as srt or ass using YOLOCR
positional arguments:
path Path to a video
optional arguments:
-h, --help show this help message and exit
-V, --version show program's version number and exit
-c CONFIG, --config CONFIG
path to configuration file
-l language, --lang language
Select the language of the subtitles (default: eng)
-wd folder, --work-dir folder
Temporary stuff directory (default ./temp)
-o folder, --output-dir folder
Output directory (default ./output)
-log level, --log-level level
Set the logging level (default INFO)
-ass style, --ass-style style
ASS style to use if sub-format is ass (default: Verdana 60)
-rr path to regex-replace json, --regex-replace path to regex-replace json
List of regex/replace for automatic correction
-hcr char,replace, --heuristic-char-replace char,replace
List of char/replace for heuristic correction
-sf format, --sub-format format
Set the outputed subtitles format (default: srt)
-m mode, --mode mode Set the processing mode.
"filter" to only start the filtering jobs
"ocr" to process already filtered videos
"full" for filter + ocr (default: full)
-vpy vpy_file, --vpy vpy_file
Vapoursynth file to use for filtering (default: extract_subs_v1.vpy)
-T number, --threads number
Number of threads the script will use (default: automatic detection)
-autosamesub number, --auto-same-sub-threshold number
Percentage of comparison to assert that two lines of subtitles
are automatically the same (default: 95)
-samesub number, --same-sub-threshold number
Percentage of comparison to assert that two lines of subtitles
are the same (default: 80)
-nsc, --no-spellcheck
Deactivate the function which tries to replace allegedly bad characters
using spellcheck (It will make the "heurist_char_replace"
option of the userconfig useless)
-t, --timid Activate timid mode
(It will ask for user input when some corrections are not automatically approved)
-d, --delay Delay correction after every video is processed
-tss path to tesseract binary, --tesseract-path path to tesseract binary
The path to call tesseract (default: tesseract)
-vps path to vspipe binary, --vapoursynth-path path to vspipe binary
The path to call vapoursynth (default: vspipe)
-wdt number, --width number
Width of the box containing the subtitles in pixels
-hgt number, --height number
Height of the box containing the subtitles in pixels
-cropy number, --cropbox_y number
Height of the subtitle box relative to the bottom in pixels
-cropalty number, --cropbox-alt_y number
Height of the alternative subtitle box (\an8) relative to the bottom in pixels.
-1 to disable (default)
-ss number, --supersampling number
Supersampling factor, -1 to disable (default)
-er number, --expand-ratio number
Expand/Inpand factor for supersampling
-rs mode, --resampler mode
Scaling algorithm to use
-wt number, --white-thresh number
Color threshold of the inner subtitles
-bt number, --black-thresh number
Color threshold of the outer subtitles (the black border)
-dt number, --detect-thresh number
General detection threshold
Lower values lead to more detected subs, more positive false
Higher values lead to subs difficult to detect
You need to specify the height (relative to the bottom) and the dimensions of the box the subtitles are contained in. So for example:
python pythOCR.py /myVideos/vid01.mp4 -l eng -sf srt --width 120 --height 150 --cropbox_y 0.
Developer for this repo:
 pandamoon21 💻 |