Logo Lanfrica

72730882/Amharic-OCR-Website

Domaine:

natural language processing

Type de record:

software
Créateur:
Hôte:
# Amharic-OCR This repository contains a backend and frontend for an Amharic OCR service. The backend exposes an endpoint that accepts image uploads, stores them in Cloudinary, and records metadata in MongoDB. ## Backend setup 1. Copy or create a `.env` file in `backend/` with the following variables: ```env MONGO_URI=" " CLOUDINARY_NAME=" " CLOUDINARY_API_KEY=" " CLOUDINARY_API_SECRET=" " ``` 2. `cd backend && npm install` 3. `node server.js` (or `npm run dev` if you have a script) – the server listens on port 5000. ## Upload endpoint **URL:** `POST localhost` - Accepts a `multipart/form-data` request - Field name for file must be `image` - Responds with JSON containing the MongoDB record and Cloudinary URL ### Example using curl ```bash curl -F "image=@/path/to/file.jpg" localhost ``` ### Example Node test script Run `node backend/test-upload.js /path/to/local.jpg` after starting the backend. This script will print the server response. --- Feel free to inspect `backend/routes/ocrRoutes.js` and `backend/controllers/ocrController.js` for the implementation.