Comic Image Text Extraction Environment

This project sets up a Dockerized development environment using Miniconda for processing comic images and extracting text via OCR.

Prerequisites

• Docker: Install Docker
• Docker Compose: (Usually included with Docker Desktop, or install separately)

Setup

1. Create Project Structure: Ensure you have the following files and directories in your project root:

   .
   ├── Dockerfile
   ├── docker-compose.yml
   ├── environment.yml
   ├── .dockerignore
   ├── README.md
   ├── notebooks/   <-- Create this directory for your Jupyter notebooks
   ├── data/        <-- Create this directory and place your comic image folders inside (not included)
   └── src/         <-- Create this directory for any custom Python modules
   

2. Place Comic Images: Put the nested directories containing your comic images inside the data/ folder.

Running the Environment

1. Build and Start:

   make build
   make run
   

2. Access Jupyter Notebook: Open your web browser and navigate to http://localhost:8888 or http://127.0.0.1:8888. You should see the Jupyter Notebook interface. Since we disabled token/password in the Dockerfile for local convenience, it should open directly.

3. Start Coding:

   • Create new notebooks inside the notebooks/ directory via the Jupyter interface.
   • Your comic images (not included) will be accessible within the container at /app/data/.
   • You can write reusable Python code in files within the src/ directory and import them into your notebooks (e.g., from src import my_module).

Stopping the Environment

make stop