RANCANG BANGUN MODEL PENGENALAN TULISAN DENGAN CONVOLUTIONAL NEURAL NETWORK (CNN)

Handwriting recognition has become a significant research topic in the fields of computer vision and natural language processing. This study aims to develop a Hijaiyah handwriting recognition system using Convolutional Neural Networks (CNN) at Taman Pendidikan Al-Qur'an (TPA) Baitul Hidayah Mosque, Kepuhrejo village, Magetan Regency. This system is expected to assist students in recognizing and memorizing Hijaiyah more effectively and support teachers in assessing students' abilities. The application development was carried out using CNN with a Tkinter-based GUI interface. The dataset used includes diverse handwritten characters to train the CNN model to recognize various handwriting styles. This application allows users to draw characters or numbers on a digital canvas, which are then recognized by the CNN model, and the recognition results are displayed in real-time. The implementation of CNN demonstrates a high accuracy of 99.19% in handwriting image classification. The testing results show that this application has a high accuracy rate and quick response time, making it an effective solution for Hijaiyah learning at TPA.