Predictive Analysis for Optimal Text Visibility: A Comprehensive Study on Frame-of-Interest Prediction in Book Digitization Videos

Articles in Press

Document Type : Original Article

Authors

1 Department of Computer Science and Engineering,Sardar Vallabhbhai National Institute of Technology,Surat,India

2 Pimpri Chinchwad College of Engineering and Technology

3 School of Computing and Data Sciences, FLAME University

Abstract

This research paper investigates the prediction of frames of interest in book digitization videos, specifically identifying frames where the entire text is optimally visible. Employing a curated dataset of diverse book flipping videos, we employ advanced machine learning techniques, including computer vision and deep learning, to develop a predictive model. The study emphasizes both technical aspects and broader implications for streamlining digitization processes. Results demonstrate the model's effectiveness, showcasing its potential to automate the identification of frames of interest, thereby enhancing the efficiency of book digitization workflows. The findings contribute to the intersection of computer vision, machine learning, and digitization efforts, offering insights for researchers and practitioners in the field. The presented model represents a significant step towards intelligent frame-of-interest prediction, with implications for optimizing the accessibility and usability of digitized textual content.We have tried to apply all the basic approaches and we are getting encouraging results with good accuracy.

Keywords

Main Subjects

International Journal of Engineering

Articles in Press, Accepted Manuscript
Available Online from 10 April 2024

Files

Share

How to cite

Statistics