Automated Classification of Arabic Script Styles Using Enhanced CNN Frameworks
Loay Alzubaidi, Mousa Sweidan · 2024
This paper presents an innovative approach for classifying Arabic Script Styles using Convolutional Neural Networks (CNNs). The goal of this project is to develop a robust classification model that accurately distinguishes between various styles of Arabic script, enhancing the understanding and recognition of this rich cultural heritage. The methodology involves feature extraction through the analysis of unique characteristics and patterns within the Arabic scripts, leveraging deep learning techniques to improve classification performance. By utilizing CNNs, the proposed system achieves impressive accuracy in identifying and categorizing different Arabic Script Styles. This research not only automates the classification process but also contributes to the preservation and appreciation of Arabic calligraphy and typography. The experimental results demonstrate the model's effectiveness, achieving a classification accuracy of 97.75%, underscoring the significant role of deep learning in the analysis and preservation of cultural artifacts.