Fig. 1: Research framework (Source: created by the authors). | npj Heritage Science

Fig. 1: Research framework (Source: created by the authors).

From: Semantic segmentation and spatial grid analysis of Chinese heritage landscape photographic compositions with cross-cultural perspectives

Fig. 1: Research framework (Source: created by the authors).

The light blue blocks represent the study cases, focusing on the Humble Administrator’s Garden (HAG) in Suzhou and the Summer Palace (SP) in Beijing. The light green blocks indicate the data collection phase, where Python-based web crawlers are used to gather UGPs from Dianping.com and TripAdvisor.com. The light orange blocks represent the data analysis phase, which includes semantic segmentation, occupancy statistics, and composition analysis. In this phase, semantic segmentation is applied using DeepLab V3+ , occupancy statistics are calculated for ten key components, and composition analysis is performed using the Rule of Thirds and k-means clustering.

Back to article page