Python Image Background Remover

An open-source, high-accuracy background removal tool powered by deep learning models. This project leverages the withoutbg and rembg engines to provide superior results for removing backgrounds from images.

Features

• High Accuracy: Utilizes advanced AI models to precisely separate foregrounds from backgrounds.
• Dual Engine Support: Choose between withoutbg (default, high accuracy) and rembg engines for your specific needs.
• Batch Processing: Process multiple images at once and download results as a ZIP file.
• Image Composition: Easily add solid colors or custom images as new backgrounds for your processed images.
• Web Interface: Simple and intuitive Django-based web interface.
• API Support: RESTful endpoints for integrating background removal into your own applications.

Technologies

• Backend: Django
• AI/ML Engines: withoutbg, rembg, onnxruntime
• Image Processing: Pillow (PIL)
• Package Management: uv / pip

Installation

Prerequisites

• Python 3.12 or higher
• uv package manager (recommended) or pip

Steps

1. Clone the Repository

   git clone https://github.com/Cyber24BD/python-image-bg-remove.git
   cd python-image-bg-remove

2. Install Dependencies

   Using uv:

   uv sync

   Or using pip:

   pip install -r requirements.txt

   (Note: If requirements.txt is not present, generate it from pyproject.toml or install dependencies manually.)

3. Run Migrations

   uv run manage.py migrate
   # or
   python manage.py migrate

4. Start the Development Server

   uv run manage.py runserver
   # or
   python manage.py runserver

5. Access the Application Open your browser and navigate to http://127.0.0.1:8000.

Usage

Web Interface

1. Go to the home page.
2. Upload an image (JPEG, PNG, WebP).
3. Select the processing engine (withoutbg for higher accuracy).
4. View the result, download it, or edit the background.

API

POST /upload/

• image: (File) The image file to process.
• engine: (String) withoutbg or rembg.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License