NESTrisOCR-CJL

Reimplementation & Simplification & Slight improvements on original NESTrisOCR for CTWC Japan Lite.

See also NESTrisSystem-CJL.

Contents

• assets/: Image files for OCR/GUI.
• captureWorker.py: A thread capturing a window region and performs OCR.
• config.py: Saves/loads configurations.
• main.py: Entry point.
• main.spec: Configuration for PyInstaller (which makes an .exe executable).
• mainWindow.py: Automatically generated from pyuic5 -o mainWindow.py mainWindow.ui. Do not edit manually.
• mainWindow.ui: Qt Designer file of the main window.
• networkWorker.py: A thread connecting and sending data to NESTrisSystem-CJL Server.
• win32.py: Win32 helper functions.

Setup

1. Install Python 3.
2. Run pip3 install numba numpy Pillow pyqt5 pywin32

Running

Run python main.py

Deploying

Run pyinstaller main.spec, then dist/NESTrisOCR/ is distributable.

What's changed from original version

• All-in-one GUI for connecting/previewing/setting
• Fixed resolution (512x448) for processing OCR
• More balanced field detection (reads center 2x2 from 4x4 block)
• Added game detection feature

What's missing compared to original version

Contributions welcome!

• macOS/Linux support
• Automatic calibration
• Different window capture methods (File, OpenCV, etc.)
• Different packet format

What I wanted to implement but had no time

Contributions welcome!

• Automatic color calibration
• Faster capturing/OCR
• Porting to web app (by using getUserMedia(), getDisplayMedia())

License

MIT. See LICENSE.