pp. 783-794
S&M3956 Research Paper of Special Issue https://doi.org/10.18494/SAM5509 Published: February 28, 2025 Change Detection of Historical Villages in the Korean Demilitarized Zone Using a CNN Model: Focusing on Villages within the Korean Demilitarized Zone during the 1910s and 1950s [PDF] Haeyong Jeong (Received December 19, 2024; Accepted February 21, 2025) Keywords: demilitarized zone, deep learning, convolutional neural network, temporal-spatial analysis, historical topographic maps
In this study, we investigated the spatial and temporal changes in village distributions within the Korean Peninsula’s Demilitarized Zone (DMZ) using deep learning methods. Historical maps from the 1910s and 1950s were analyzed to analyze house distributions and identify changes caused by historical events. A custom convolutional neural network model was developed for automated feature extraction, achieving high accuracy compared with traditional methods. The findings provide foundational data for understanding the historical continuity of settlements within the DMZ and aim to support future research on its restoration and development.
Corresponding author: Haeyong Jeong![]() ![]() This work is licensed under a Creative Commons Attribution 4.0 International License. Cite this article Haeyong Jeong, Change Detection of Historical Villages in the Korean Demilitarized Zone Using a CNN Model: Focusing on Villages within the Korean Demilitarized Zone during the 1910s and 1950s, Sens. Mater., Vol. 37, No. 2, 2025, p. 783-794. |