INTELLIGENT GRAMMER REWRITE USING OCR AND NLP TECHNIQUES

Authors:

1. P Kamakshi¹, G. Sai Saranya², CH. Suvarna Raju ³, B. Jahnavi⁴, M. Manoj Kumar⁵, A. Sree Ratna Teja⁶,
2. B Hari Priya⁷

^1234567Department of Information Technology
Dhanekula Institute of Engineering and Technology

Abstract - Handwritten documents are still common in schools, offices, and daily personal tasks, even with digital technologies available. However, turning handwritten text into an editable digital format is challenging. Handwriting styles differ from person to person, spacing can be irregular, and images may have noise or poor lighting. Here, paper introduces a smart system for recognizing the handwritten text images and improving spelling and grammar by combining transformer-based Optical Character Recognition with Natural Language Processing techniques. After applying suitable preprocessing techniques, the system analyses handwritten documents line by line using the TrOCR model to accurately recognize text. In addition to text extraction, the system improves usability by providing features such as synonym suggestions, grammar and spelling correction, and text-to-speech functionality. A simple and interactive graphical user interface is developed using Python Tkinter to ensure ease of use for all users. Since the system functions entirely in offline mode, it allows access without an internet connection while maintaining complete data privacy. So here, the proposed system provides usable and reliable approach for converting handwritten text images into digital format, which can be more used in professional, academic and documentation purpose work.

Keywords— Handwritten Text Recognition, TrOCR, Natural Language Processing, Optical Character Recognition, Grammar and Spelling Correction.