Claim Missing Document
Check
Articles

Found 2 Documents
Search

Classifying Hijaiyah Letters Handwritten Detection of Children Using CNN Algorithm Roofiad, Ahmad Maulidi; Sabillah, Annisa; Rohman, Aprian Nur; Triyani, Elmi Wahyu
Khazanah Journal of Religion and Technology Vol. 2 No. 1 (2024): June
Publisher : UIN Sunan Gunung Djati Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15575/kjrt.v2i1.812

Abstract

Learning the Hijaiyah letters is an important basis because in learning the Qur'an, these abilities must be mastered before they can be introduced and taught to children. However, the recognition of Hijaiyah letters in children's handwriting is still a challenge due to the variations and inconsistencies that are often found. Deep learning technology, particularly Convolutional Neural Network (CNN), has demonstrated its ability to classify letters with a high degree of accuracy. Therefore, this research aims to develop a CNN-based Hijaiyah letter classification model to help children learn to write and read Hijaiyah letters properly. This research uses a CNN model that is optimized with data augmentation techniques and hyperparameter tuning. The model was trained using a standard dataset totaling 1,740 samples of Hijaiyah letters. Model evaluation is done by calculating accuracy, precision, recall, and F1-Score on the validation dataset. The results showed that the proposed CNN model achieved almost 94.35% accuracy on the validation dataset. This research is expected to improve children's ability to learn Hijaiyah letters.
Klasifikasi Tulisan Tangan Huruf Hijaiyah Anak Usia 6-8 Tahun Menggunakan Metode Support Vector Machine Roofiad, Ahmad Maulidi; Alam, Cecep Nurul; Atmadja, Aldy Rialdy
SENTRI: Jurnal Riset Ilmiah Vol. 4 No. 12 (2025): SENTRI : Jurnal Riset Ilmiah, Desember 2025
Publisher : LPPM Institut Pendidikan Nusantara Global

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55681/sentri.v4i12.5077

Abstract

This study aims to develop a handwritten Hijaiyah letter classification system for children aged 6–8 years using the Support Vector Machine (SVM) algorithm. The main problem in elementary education is the difficulty children face in recognizing and writing Hijaiyah letters due to the similarity of their shapes and variations in handwriting. The research process uses the CRISP-DM stages, consisting of problem understanding, data collection and preparation, modeling with SVM (GridSearch for hyperparameter tuning), and evaluation using a confusion matrix and f1-score. The dataset used consists of 2,100 images of handwritten letters from elementary school students. The results show that the SVM model with RBF kernel, C=10, and gamma="scale" achieved the highest accuracy of 83.57%. This study demonstrates that an SVM-based machine learning approach can assist in recognizing Hijaiyah letters, making it a practical solution for teachers in teaching Hijaiyah writing.