🧠 Project Title: Capture – Text Extraction from Image Captchas 📌 Overview: "Capture" is a web application designed to extract text from image-based captchas using deep learning. The application allows users to upload captcha-style images, and it responds with the detected text. It's a demonstration of integrating deep learning with web development for real-time image processing.
💻 Tech Stack: 🔹 Frontend (React.js):
Built with React for a dynamic, responsive interface. Four key sections: Home: Hero section with a catchy heading and illustrative image. About: Overview of the project’s purpose and usage. Model Specs: Description of the ML/DL model used for text extraction. Capture: Main interface to upload captcha images and display the predicted text. 🔹 Backend (Flask):
Handles API endpoints for image uploads and returns predictions. Integrates the trained deep learning model for text recognition. 🔹 Deep Learning (Python):
Model trained to recognize alphanumeric captcha characters. Trained on a synthetic dataset CaptchaMNIST.
.
