Gazou is a Japanese OCR for Linux written in C++. It utilizes image processing, and other parameters to improve the accuracy. All contributions are welcome.
First, enable my gentoo overlay with eselect-repository
. If you don't have it already, follow the guide on the Gentoo wiki first.
eselect repository add kamui git https://github.com/kamui-fin/gentoo-overlay.git
emerge --sync kamui
Then you can emerge the package normally with:
emerge -av gazou
Since this is packaged in the AUR, you can install it with any AUR helper like:
paru -S gazou-git
These dependencies will need to be installed by your system's package manager:
- Qt5 >= 5.10
- Tesseract >= 4.0.0
- Leptonica >= 1.70
Optional dependencies:
- Qt5X11Extras >= 5.10 (for GUI)
git clone --recursive https://github.com/kamui-fin/gazou.git
cd gazou
mkdir build
cd build
cmake .. -DGUI=ON
sudo make install
For wayland users, gazou
must be compiled with -DGUI=OFF
and use the equally functional command line interface.
This is due to the fact that Wayland does not support applications registering their own global hotkeys as a security measure.
You can run the program by typing gazou
in your terminal. It runs in the background and should appear in your system tray.
If you click on the app in the tray, a settings dialog should appear.
In here, you can customize the keybinds to your liking. The default hotkeys are:
Alt+A
: Vertical OCRAlt+D
: Horizontal OCRAlt+S
: Repeat the previous OCR
When you perform an OCR, the result text gets copied to the clipboard.
Gazou also has a command line mode, and this can be useful for integrating it with bash scripts.
To get the resulting text copied to your clipboard, you can use xclip
, wl-copy
, or any clipboard utility you prefer.
Available options include:
1. gazou
Run the main application
2. gazou --help
Display this message
3. gazou prevscan
Run the OCR on the same coordinates of the previous scan
4. gazou ORIENTATION{-h; -v; horizontal; vertical}
Interactively run the OCR and print the output to stdout
5. gazou ORIENTATION{-h; -v; horizontal; vertical} IMAGEFILE
Run the OCR on the given IMAGEFILE with the given ORIENTATION.