Azure OCR API Wrapper

By: Andreas Lambropoulos

This is a simple API Wrapper for Optical Character Recognition (OCR) using Microsoft’s Azure Optical Character Recognition service. It allows the user to input an image and it returns the extracted text.

API Access and Documentation

• Base URL: http://147.182.167.158:3000/api/ocr
• Swagger UI: http://147.182.167.158:3000/docs/
• User Guide: user_guide.md

Key Features

• Extracts text from an image using Azure OCR service
• Supports both image URLs and image uploads with POST /url and POST /upload endpoints
• Validates file format and size before processing
• Returns clean and readable JSON responses with status codes
• Swagger UI available for request/response formats and live testing
• Deployed to a DigitalOcean droplet with PM2 server

Image Requirements

• Supported formats: All image formats, including JPG, PNG, BMP, TIFF, WEBP, AVIF, and even GIF
• Max file size: 4 MB
• Dimensions: Images must be between 50 and 4,200 pixels in both width and height

Tech Stack and Implementation

• Node.js + Express – Backend server framework for handling API endpoints
• Axios – Sends HTTP requests to Azure OCR's REST API with proper headers and body
• CORS – Enables cross-origin access for API testing
• Multer – Handles file uploads with multipart/form-data and enforces the 4MB size limit
• Sharp – Converts unsupported image formats (like WebP or AVIF) to PNG to ensure compatibility with Azure OCR
• Swagger (swagger-jsdoc + swagger-ui-express) – Automatic generator for interactive API documentation
• DigitalOcean Droplet – Continuously hosts the Express server with PM2

Image Sources

You can find these images in the test_data folder of this repository, which were obtained from:

• MSRA-TD500 Text Detection Dataset
• Other custom image searches and manually created examples

Build Your Own Local OCR API

1. Clone the repository:
   git clone https://github.com/andlamb2002/ITIS-6177-API-Wrapper.git

2. Install dependencies:
   npm install

3. Create an Azure Computer Vision resource through the Azure Portal. Then, add a .env file in the root directory with your Azure credentials:

   AZURE_API_KEY=your_api_key
   AZURE_ENDPOINT=https://your-azure-endpoint/

4. Start the server:

   node server.js

5. Test the API endpoints:

• POST http://localhost:3000/api/ocr/url – Submit an image URL in a JSON body using the imageUrl field
• POST http://localhost:3000/api/ocr/upload – Upload an image file using multipart/form-data with the field name file

Or access the Swagger UI at: http://localhost:3000/docs

Disclaimer

This project was created as part of the ITIS-6177 System Integration course at UNC Charlotte. It is intended for educational and demonstration purposes only.