Build a Text Extractor App with Python Code Under 30 Lines Using Gradio and Hugging Face

original post: https://baxin.netlify.app/build-text-extractor-python-under-30-lines/

Extracting text from images, known as Optical Character Recognition (OCR), is a valuable feature for applications in document processing, data extraction, and accessibility. In this guide, we will create an OCR app using Python libraries like pytesseract for OCR, Pillow for image processing, and Gradio for building an interactive UI. We’ll deploy this app on Hugging Face Spaces.

Prerequisites

Before starting, you’ll need a Hugging Face account and basic familiarity with Docker.

Step-by-Step Guide

Step 1: Create a Hugging Face Space

1. Navigate to Hugging Face Spaces: Log in to Hugging Face and go to the "Spaces" section.
2. Create a New Space:
   • Click on "New Space."
   • Name your space (e.g., image-text-extractor).
   • Choose Gradio as the SDK and set the visibility (public or private).
   • Click "Create Space."

Step 2: Create a Dockerfile

To deploy on Hugging Face Spaces with required system dependencies, such as Tesseract for OCR, we need a Dockerfile that configures the environment.

Create a Dockerfile with the following content:

# Use an official Python runtime as a parent image
FROM python:3.12
ENV PIP_ROOT_USER_ACTION=ignore

# Set the working directory in the container
WORKDIR $HOME/app

# Install system dependencies
RUN apt-get update && apt-get install -y
RUN apt-get install -y tesseract-ocr
RUN apt-get install -y libtesseract-dev
RUN apt-get install -y libgl1-mesa-glx
RUN apt-get install -y libglib2.0-0
RUN pip install --upgrade pip

# Copy requirements and install dependencies
COPY requirements.txt requirements.txt
RUN pip install --no-cache-dir -r requirements.txt

# Copy the app code
COPY app.py ./

# Expose the port for Gradio
EXPOSE 7860

# Run the application
CMD ["python", "app.py"]

Step 3: Create the OCR Application

1. Create a file called app.py with the following content:

import gradio as gr
import pytesseract
from PIL import Image
import os

def extract_text(image_path):
    if not image_path:
        return "No image uploaded. Please upload an image."

    if not os.path.exists(image_path):
        return f"Error: File not found at {image_path}"

    try:
        img = Image.open(image_path)
        text = pytesseract.image_to_string(img)
        return text if text.strip() else "No text detected in the image."
    except Exception as e:
        return f"An error occurred: {str(e)}"

iface = gr.Interface(
    fn=extract_text,
    inputs=gr.Image(type="filepath", label="Upload an image"),
    outputs=gr.Textbox(label="Extracted Text"),
    title="Image Text Extractor",
    description="Upload an image and extract text from it using OCR."
)

iface.launch(server_name="0.0.0.0", server_port=7860)

1. Create a requirements.txt file to specify the dependencies:

gradio
pytesseract
Pillow

This setup includes:

• Image Upload: gr.Image(type="filepath") allows users to upload images as file paths, which pytesseract processes.
• Text Extraction: pytesseract.image_to_string extracts text from the image.
• User Interface: Gradio generates a simple UI for users to upload an image and view extracted text.

Step 4: Push All Files to Hugging Face Spaces

With all files created, push them to your Hugging Face Space