Handwriting Recognition And Information Retrieval System

Authors: Assistant Professor Dr. Madhusudan Kulkarni, Mr. Abdulrahman Shaikh, Ms. Ankita A Chavan, Mr. Atiqur Rehman Sayed, Mr. Rahul Madiwal

Abstract: This project is a full-stack handwriting recognition app that blends machine learning with modern web technologies to turn handwritten text into digital content. The frontend is built using React and JavaScript, while the backend runs on Python and Django. At its core, the app uses a pre-trained machine learning model based on the EMNIST dataset to recognize handwritten digits, uppercase letters, and unique lowercase letters. To tell similar characters apart, it even uses a clever height-based approach to accurately detect letter casing.Designed with real-world use in mind—like in education, healthcare, and business—the app helps quickly digitize handwritten notes and forms. It also features ChatGPT integration, so users can get instant explanations and context for the text they've scanned, making the experience more interactive and insightful.By combining deep learning with AI, the app delivers accurate handwriting recognition, quick info retrieval, and a clean, user-friendly interface. It’s a powerful tool for boosting productivity and simplifying the move from paper to digital.