Warning
This software isnt supported, maintained, tested or is deprecated since 2023.
Important
This software won't run on its own. All the referenced media template matching files in the source code are cleaned. This repository serves only educational purposes.
This is a small automation script, to demonstrate the usecase of image recognition in automatization of video games for educational purposes.
Warning
This version supports 2160p and 1080p screens in fullscreen or windowed-left-top-cornered mode (recommended). It does not support other resolutions or windowed positions because the coordinates for screen windows (in which template matching occurs) are not retrieved dynamically or by percent yet.
Fisher.py is the main file, which contains the class Fisher. HoleFisher.py contains a semi-automated version, where the user can fish without the effort to play the minigame.
Install all dependencies in your Python environment with pip from the root of the repository directory.
pip install .
Run
python starter.py [-f] [-s] [-a]
-f : fullscreen mode
-s : semi-auto mode
-a : automatic mode
Make sure to fulfill the correct UI scaling option and resolution in its settings.
Bindings
- Fishing on Hotkey 'j'.
- Weapon Switch on Circumflex '^'.
- Weapon Ability Bar Default Num Keys 1-5.
- Dock your Window in the top left corner of your Windows Desktop.
Warning The docking ensures that every part of the screen is correctly matched in pixels. Without it, the automatic mode won't work.
- Place your character in front of the water.
- Check every few hours, there can be false positives for the exclamation mark check because of the handcrafted threshold value.
Follow the command line instructions after starting.