OCR Receipt Total Extractor

A FastAPI-based web application that uses OCR (Optical Character Recognition) to automatically extract totals from receipt and invoice images. The system is containerized with Docker for simple and fast deployment.

Features

• Intelligent OCR: Smart total extraction from receipt images using advanced pattern recognition
• Multi-format Support: Supports all common image formats (JPEG, PNG, WEBP, BMP, TIFF)
• Web Interface: easy image uploads
• REST API: RESTful endpoints for programmatic integrations
• Multi-language: Recognizes totals in Italian, English, French, German, and Spanish
• Data Persistence: Automatic saving of results in JSON format
• Containerized: Docker and Docker Compose

Prerequisites

• Docker and Docker Compose installed on your system
• At least 2GB of RAM available (for OCR models)
• Port 8000 available
• Modern web browser (Chrome, Firefox, Safari, Edge)

Quick Start

Method 1: Docker Compose (Recommended)

1. Clone the repository: git clone cd ocr-receipt-extractor

2. Avvio applicativo Docker

• docker compose up -d
• docker compose ps

Metodo 2: Local Installation

• pip install -r requirements.txt
• uvicorn main:app --host 0.0.0.0 --port 8000

PROJECT STRUCTURE

ocr-receipt-extractor/ ├── main.py
├── requirements.txt
├── Dockerfile
├── docker-compose.yml
├── templates/ │ └── index.html
├── static/
├── data/ │ ├── media/
│ └── data.json
└── README.md