99sbr / OCR-Service-Restapi

OCR-Service-Restapi

Overview

A FastAPI-based simple web service that accepts images, runs optical character recognition (OCR) on them and returns the extracted text.

Prerequisites

Before you begin, make sure you have the following prerequisites installed:

• Python 3.10
• Pip (Python package installer)
• Docker
• fastapi
• uvicorn
• python-dotenv
• pyyaml
• python-multipart
• pytesseract
• Pillow

Installation

1. Clone the repository:

   git clone https://github.com/99sbr/OCR-Service-Restapi.git

2. Navigate to the project directory:

   cd OCR-Service-Restapi

Containerize the Service

To run the web service in a Docker container:

1. Build the Docker image:

    docker build -t ocrservice-image  .

2. Run the Docker container:

    docker run -p 80:80 ocrservice-image

Visit http://localhost:80/docs to access the Swagger documentation and interact with the API.

API Endpoints

1. POST /imgsync
   • Description: Synchronously extracts text from an image.
   • Request Body: Provide a JSON object with the base64-encoded image data.
   • Response: Returns a JSON object with the extracted text.
2. POST /imgasync
   • Description: Asynchronously extracts text from an image.
   • Request Body: Provide a JSON object with the base64-encoded image data.
   • Response: Returns a JSON object with a job_id for tracking the asynchronous task.
3. GET /ocr_text
   • Description: Asynchronously gets OCR text from Job ID.
   • Request Body: Provide a JSON object with the job-ID data.
   • Response: Returns a JSON object with the extracted text.

Code Structure

• api_template: Contains all the API-related Code Base.

  • manage.py: Only entry point for API. Contains no logic.
  • .env: Most important file for our API and contains global configs. Avoid using application/variable level configs here.
  • application: It contains all our API-related codes and test modules. I prefer keeping the application folder global.
  • logs: Logs are self-explanatory. FYI it will not contain any configuration information, just raw logs. Feel free to move according to your comfort but not inside the application folder.
  • settings: Logger/DataBase/Model global settings files in yaml/json format.

• application:

  • main: priority folder of all your application-related code.
    • 📮 routers: API routers and they strictly do not contain any business logic
    • 📡 services: All processing and business logic for routers here at the service layer
    • ⚒ utility:
      • logger Logging module for application
      • manager A manager utility contains workers and handlers for Data Related Tasks which can be common for different services.
  • initializer.py: Preload/Initialisation of Models and Module common across applications. The preloading model improves inferencing.