ComfyUI-Easy-DotsOCR

ComfyUI-Easy-DotsOCR is a custom node for ComfyUI that provides text extraction via the DotsOCR engine. This node enables users to extract text from images directly within ComfyUI workflows.

Installation

1. Clone this repository into your ComfyUI custom_nodes directory:

   git clone https://github.com/yolain/ComfyUI-Easy-DotsOCR
   

2. Install required dependencies:

   pip install -r requirements.txt
   

3. Restart ComfyUI. The DotsOCR node will be available in the node list.

Usage

• Add the DotsOCR node to your ComfyUI workflow.
• Connect an image input to the node.
• Run the workflow to extract text from the image.

Download Models (Manual)

Download All models to ComfyUI/models/LLM/dots-ocr

Development

This project is developed based on dots.ocr. Many core OCR functionalities are adapted from this upstream project. Please refer to their repository for more details and credits.

File Structure

• dots.py: Core OCR logic
• nodes.py: ComfyUI node definitions
• utility/: Helper functions for image and download operations
• prompt_templates.yaml: Prompt templates for OCR tasks
• requirements.txt: Python dependencies

License

This project inherits the license from dots.ocr. Please review their license terms before use.

Acknowledgements

• dots.ocr for the original OCR implementation
• ComfyUI for the node workflow framework

Contact

For issues or feature requests, please open an issue on this repository.