Intelligent mapping of urban scenic view corridors based on public perception: A case study of Chongqing, China

Source abstract

The rapid urbanization process has increased the need to protect the scenic view corridors associated with iconic urban landscapes. However, existing approaches primarily rely on static viewshed analysis and expert judgement, which are limited to capture dynamic public movement and perception preferences. Therefore, this study proposes an intelligent framework that uses multi-source data, a random walk model, and a random forest algorithm to identify and classify urban scenic view corridors based on public perception. Using Wanzhou District as a case study, the researchers identified 105 stable viewpoints from 443,309 candidate locations and classified five types of scenic view corridors, revealing significant spatial heterogeneity. The proposed model achieved an overall accuracy of 0.82 ± 0.05, a macro-average F1 score of 0.78, and a Kappa coefficient of 0.75, indicating strong classification performance and robustness. Compared with traditional viewshed- and rule-based methods, the proposed framework demonstrates improved accuracy and interpretability by incorporating behavioral dynamics into spatial analysis. This study provides a methodological reference for identifying scenic view corridors in cities with complex terrain and supports experience-oriented urban design and visual environment management.