Density-based Clustering Algorithm for the Protection and Planning of Ancient Villages

Heritage conservation and urban planning of ancient villages demand data-driven methods. This paper investigates a clustering algorithm-based model for spatial and cultural feature analysis of historical villages, with case studies of Xidi and Hongcun. Existing methods are based on conventional geographic mapping and heuristic planning, which tend to ignore intricate socio-environmental interactions, resulting in inefficient use of resources and ineffective heritage conservation. Traditional clustering algorithms such as DBSCAN, k-means, and hierarchical clustering are limited by noise insensitivity, non-scalability, and rigidity in accommodating varied village configurations. To overcome these limitations, we introduce a sophisticated density-based clustering algorithm that combines DBSCAN with an optimized density threshold and deep feature extraction. The model improves spatial segmentation and correctly classifies areas according to architectural coherence, population density, and cultural importance. It surpasses conventional approaches, with 96.4% accuracy and a 94.8% F1-score. Comparative analysis emphasizes its excellence in spatial coherence preservation and predictive reliability enhancement. The algorithm provides a sustainable, scalable, and automated decision-support system, enabling policymakers to make informed decisions based on data-driven insights for efficient heritage preservation and strategic planning of ancient villages.