LiveTranslator

LiveTranslator finds and translates texts from images within the region or window using OCR and overlays the translated texts.

Getting Started

Requirements

• Python, Pip, Pipenv

Setup Instructions

1. Clone the repository:

   git clone <repository-url>
   cd LiveTranslator

2. Run the environment setup script:

   python setup_env.py

3. Install the dependencies:

   pipenv install

4. Use the virtual environment:

   pipenv shell

5. Exit/Deactivate the virtual environment:

   exit

Usage

Translate Text from Images

Use the following command to translate text from images:

python3 scripts/main.py translate --src <path_to_image> --print --print-boxes --from <lang> --to <lang> --translator <translator>

CLI Options Explanation

• --src (Required): Specify the path to the image file to be translated.
• --print (Optional): Print extracted texts to the console.
• --print-boxes (Optional): Draw bounding boxes around the detected text in the image.
• --from (Optional): Specify the source language for OCR (e.g., "eng", "jpn"). If not specified, it's auto-detected.
• --to (Optional): Specify the target language for translation (default is "eng").
• --show (Optional): Show the final translated image.
• --translator (Optional): Choose the translation service (default is "google"). Options: google, deepl.

Run the GUI

Use the following command to run the application's GUI:

python3 scripts/main.py gui

GUI Features

The LiveTranslator GUI allows users to capture and translate text from specific windows or regions of their screen. The main features include:

1. Select Capture Mode:

   • Capture in Intervals: Capture the selected window/region at regular intervals.
   • Capture and Wait: Capture the selected window/region once and wait until the translated window is closed.

2. Set Capture Interval:

   • Set the interval (in seconds) for capturing the screen in Capture in Intervals mode.

3. Select Source and Target Languages:

   • Choose from supported languages (e.g., English, Japanese, French) for OCR and translation.

4. Select Capture Type:

   • Window: Capture a specific window.
   • Region: Capture a user-defined region of the screen.

5. Window List:

   • Refresh and select from a list of currently open windows for capturing.

6. Region Selection:

   • Manually select a region of the screen for capturing and translation.

7. Start and Stop Capture:

   • Start capturing the selected window/region and translating the text.
   • Stop the capture process.

8. Status Display:

   • A status label displays the current state of the capture process, with tooltips for additional information.

Testing

Use the following command to run tests:

python3 tests/run_tests.py <modules>

CLI Options Explanation

• modules (Optional): Paths to specific test modules or classes. Leaving empty will run all tests
• -v, --verbosity (Optional): Verbosity level for test output.
  • 0: Minimal output
  • 1: Normal output (default)
  • 2: Detailed output
  • 3: Debug-level output
• --failfast (Optional): Stop running tests on the first failure.
• --buffer (Optional): Buffer stdout and stderr during test execution.

Troubleshooting

• If you encounter issues, you can remove the virtual environment and reinstall dependencies:

  pipenv --rm
  pipenv --python 3.9
  pipenv install

Example Usage

To translate text from an image using the Google Translator, run:

python3 scripts/main.py translate --src path/to/your/image.png --print --print-boxes --from eng --to jpn --translator google --show

Running the Application

To run the application GUI, use:

python3 scripts/main.py gui