Abstract
This study proposes an AI-based license plate recognition system deployed at parking areas in Shirakawa-go, a UNESCO World Heritage site, to analyse and visualize tourist dynamics. The system automatically detects and recognizes vehicle license plates from camera footage, enabling large-scale and continuous monitoring of tourist traffic patterns. The extracted data provide insights into visitor origins, stay durations, visit frequencies, and rental car usage, offering valuable indicators for regional tourism management. Compared with conventional manual counting and questionnaire surveys, the proposed approach enables efficient, automated, and continuous data collection suitable for small-scale destinations with limited resources. The empirical operation in Shirakawa-go demonstrated that the system can identify temporal peaks in parking demand and support cloud-based data aggregation for real-time analysis. These findings highlight the system’s potential for overtourism mitigation, parking congestion alleviation, and data-driven resource optimization, thereby contributing to sustainable and responsible tourism management in culturally significant rural areas.
