An integrated framework to quantify tour-view experience in classical gardens using computer vision and spatial analysis

Classical Chinese gardens center on the “tour-view experience” (游观), yet quantitative approaches to interpret its unfolding remain scarce. Using Ji Xing Garden (New York) as a case, we integrate space syntax, computer vision, landscape metrics, and a subjective rating experiment to build a unified “spatial structure–visual content–visual perception–dynamic sequence” framework. Space syntax characterizes the spatial skeleton of movement and visibility; visual focus and accessibility reveal how viewpoint shifts and route hierarchy guide touring. Visual morphology indices quantify element composition, form complexity, and fragmentation, and track their fluctuations along the walking sequence. Pearson correlations link objective metrics to perceived richness, tranquility, beauty, openness, and exploration desire, supporting perceptual validity. The results reinterpret key garden-making principles and show that well-calibrated visual transitions and hierarchical spatial organization strengthen experiential layering and rhythm. The workflow provides a reproducible paradigm and design reference for overseas reinterpretation of classical Chinese gardens.