ScreenshotOCR

ScreenshotOCR is a Python project that extracts text from images captured from the clipboard and processes them using ChatGPT API. It supports both regular and translated text extraction.

Openai API key is required

Features

• Capture images directly from the clipboard.
• Auto-translate functionality using prompts defined in prompt.json.

Setup and Usage

1. Create conda environment and install required dependencies:

   conda create -n sOCR python==3.9
   

   conda activate sOCR
   

   pip install -r requirements.txt
   

2. Run python main.py
3. Configure your API key and other settings using the UI.
4. Use Windows snipping tool (win+shift+s) to clip an image.
5. Use the defined hotkey (ctrl+alt+x) to trigger OCR process
6. The extracted (and optionally translated) text will be copied to the clipboard.

Customization

• Modify prompt.json to adjust the prompts for translation and output.