Deep Learning And Feature Detection Integration For Dating Heritage Architecture: A Case Study On Kantajew Temple
Journal of Bangladesh History and Archaeology · 16.02.2026 · 2 · 175-196
DOI: 10.64853/jbha20260210 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Bangladesh History and Archaeology, online erschienen am 16.01.2026. Verfasst von Md. Samaun Hasan und 3 weiteren Autor:innen.
Abstract (Originalsprache)
In the integration of Convolutional Neural Networks (CNNs) with advanced image processing offers a powerful approach for objective and real-time heritage architecture analysis. Traditional methods for dating monuments rely on manual observation, often resulting in subjective and inconsistent outcomes. This study applies a computational framework to the Kantajew Temple in Dinajpur, Bangladesh, combining CNNs with feature detection techniques—such as edge, contour, line, and corner extraction—to analyze decorative motifs, spatial arrangements, and structural patterns. The system was trained and validated to classify architectural characteristics representative of the Sultanate and Mughal periods. Achieving a classification accuracy of 97.24%, the model demonstrates strong potential for data-driven heritage analysis. This research underscores the role of AI in bridging computational vision and cultural interpretation, providing a sustainable, scalable, and precise methodology for documenting, preserving, and studying South Asian architectural heritage while enhancing both scholarly research and educational outreach.
Zitieren
APA
Hasan, M. S., Khatun, Z., Arju, H. A., & Ali, M. (2026). Deep Learning And Feature Detection Integration For Dating Heritage Architecture: A Case Study On Kantajew Temple. Journal of Bangladesh History and Archaeology, 2, 175-196. https://doi.org/10.64853/jbha20260210
Vancouver
Hasan MS, Khatun Z, Arju HA, Ali M. Deep Learning And Feature Detection Integration For Dating Heritage Architecture: A Case Study On Kantajew Temple. Journal of Bangladesh History and Archaeology. 2026;2:175-196. doi:10.64853/jbha20260210
BibTeX
@article{Hasan2026,
author = {Hasan, Md. Samaun and Khatun, Zahura and Arju, Hosne Ara and Ali, Mohammad},
title = {Deep Learning And Feature Detection Integration For Dating Heritage Architecture: A Case Study On Kantajew Temple},
journal = {Journal of Bangladesh History and Archaeology},
year = {2026},
volume = {2},
pages = {175--196},
doi = {10.64853/jbha20260210}
}
EUVIMED