AI-BASED PDF READER FOR REAL-TIME PROCESSING TO ASSIST BLIND STUDENTS

ABSTRACT In the modern educational landscape, access to digital resources is critical for academic success. However, students with visual impairments face significant barriers when interacting with digital documents such as PDF files, which are often designed for visual consumption. This project addresses this challenge by developing a PDF Reader for Blind Students—an innovative, assistive software solution that transforms static PDF documents into dynamic, accessible audio content. The primary objective of this project is to provide visually impaired students with an inclusive tool to independently access, navigate, and comprehend textual information embedded in PDF files. The system leverages Optical Character Recognition (OCR) to extract text from image based PDFs, ensuring compatibility with scanned documents and handwritten materials. Advanced Natural Language Processing (NLP) techniques are employed to analyze the text structure, enabling features such as heading detection, paragraph segmentation, and context aware reading. Once the content is extracted, the system uses a Text-to-Speech (TTS) engine to convert the text into natural, human-like speech. Users can control playback speed, choose from different voice profiles, and navigate the document by page, heading, or section, providing a flexible and user centric experience. I.INTRODUCTION In the rapidly evolving landscape of digital education, resources such as textbooks, research papers, and lecture notes are increasingly distributed in PDF format due to their portability and universal compatibility. This shift has streamlined access for many students, enabling seamless sharing and storage of academic content. However, for blind or visually impaired students, this transition presents significant challenges. Traditional assistive technologies, such as braille displays and tactile graphics, are often inadequate for digital formats, particularly PDFs, which may include complex layouts, embedded images, or multi-column structures. Screen readers, while widely used, frequently struggle to interpret these elements accurately, leading to fragmented or incomplete readings that hinder comprehension