A very simple tool for removing on-screen text from video, using a user-provided mask
Project description
cleancredits
A very simple tool for removing on-screen text from video, using a user-provided mask.
Dependencies
cleancredits has the following system dependencies:
Installation
-
Mac OS: Install homebrew and run
brew install python-tk
to install tkinter bindings. -
Install pipx, a package manager for python tools.
-
Install the latest released version of cleancredits:
pipx install cleancredits
Or install the latest version from Github:
pipx install git+https://github.com/BeatriceEagle/cleancredits@main
Usage
Generate HSV mask
cleancredits mask [OPTIONS] VIDEO
Generate a mask based on a video clip
Options:
-s, --start TIMECODE Start timecode (HH:MM:SS[:frame]) in the input
video
-e, --end TIMECODE End timecode (HH:MM:SS[:frame]) in the input video
-i, --input FILE Input mask. These pixels will always be present in
the output mask (unless explicitly excluded).
-o, --output FILE Output mask to this location [required]
--hue-min INTEGER RANGE Minimum hue [0<=x<=179]
--hue-max INTEGER RANGE Maximum hue [0<=x<=179]
--sat-min INTEGER RANGE Minimum saturation [0<=x<=255]
--sat-max INTEGER RANGE Maximum saturation [0<=x<=255]
--val-min INTEGER RANGE Minimum value [0<=x<=255]
--val-max INTEGER RANGE Maximum value [0<=x<=255]
--grow INTEGER RANGE Grow amount [0<=x<=20]
--bbox-x1 INTEGER RANGE Bounding box left x [x>=0]
--bbox-x2 INTEGER RANGE Bounding box right x [x>=0]
--bbox-y1 INTEGER RANGE Bounding box top y [x>=0]
--bbox-y2 INTEGER RANGE Bounding box bottom y [x>=0]
--gui / --no-gui Set --no-gui to directly render the mask without
displaying the GUI
--help Show this message and exit.
This command will display a graphical interface for modifying a mask that allows isolating part of an image based on hue / saturation / value, as well as a bounding box. You can also manually add or exclude parts of an image.
You can layer combine masks for multiple colors or areas of credits by outputting a mask, then passing that as an --input
to the mask
command.
Clean credits
cleancredits clean [OPTIONS] VIDEO MASK
Arguments:
VIDEO
: Path to the video file being cleanedMASK
: Path to the mask file. This should be a black and white png, where white indicates areas that will be removed and interpolated.
Options:
-
--start/--end
: The start/end timecodes (HH:MM:SS[:frame]) to process from the input video. Default: Start and end of the input video. -
--radius
: The number of pixels the inpaint algorithm uses for interpolation. The default is 3, and this generally gives good results, but if you want to experiment, go wild. -
--framerate
: The framerate (fps) of the video being cleaned. The default is the input framerate. -
--output PATH
: If this flag is selected, the cleaned frames will be remuxed into video and output at the specifiedPATH
. You can omit this option if you want to do your own muxing.cleancredits
muxes video using ffmpeg's libx264 codec and yuv420p colorspace, which in testing were found to give the best quality video while also still being recognizable by most editors and players. Outputting as a.mp4
file is recommended.
Example:
# Takes a video.mkv and mask.png from the current directory and outputs
# output.mp4 to the current directory.
cleancredits clean video.mkv mask.png -o output.mp4
Output:
A folder containing all the cleaned frames of the video, and the cleaned video.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file cleancredits-2.0.0.tar.gz
.
File metadata
- Download URL: cleancredits-2.0.0.tar.gz
- Upload date:
- Size: 27.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.11
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | e45177b4e74e5a6dca646988570b9632aaf9df0a2135929b50a4e095d440cb6d |
|
MD5 | a4fd1aa602106d35e5b1717d8dc24751 |
|
BLAKE2b-256 | 6615c0eba68cc5e82e84bad99faf4fe389d8a8fcd9b9e700fab22c944e3fa6ec |
File details
Details for the file cleancredits-2.0.0-py3-none-any.whl
.
File metadata
- Download URL: cleancredits-2.0.0-py3-none-any.whl
- Upload date:
- Size: 26.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.11
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d84551958efcd6fb44fdf496d57d0021a790a0443665296244b933830d278a82 |
|
MD5 | 702019da4b63821b1d31089cac3b31e8 |
|
BLAKE2b-256 | e072d2a10bed2f4c3dd5fa49538595391373455d450bdf4a5d1ab325033b3af6 |