Association Between Landscape Features Of Park Trail And Brisk Walking Preference: Impact And Design Guidance

PURPOSE: The purpose of this study was to explore the relationship between the landscape features of park trails and brisk walking preference. This paper analyzes how the landscape features of park trail affect the choice of walking path. METHODS: By crawling the street view data of Shanghai and Nanjing parks, machine learning was used to identify the image features. And the data of walking record were obtained from Strava website to identify the high-heat path. Ordinary least squares (OLS) and spatial error models (SEM) were used to analyze the coupling strength between a single feature and a combination of features and the high-eat path. Based on the combination of features, using Java language to develop a route selection model, the park trail landscape design scheme is organized on 3 km, 5 km and 10 km trails. RESULTS: The analysis results showed that the coupling strength between the tree features and the high-heat path was the highest, and the marker was the lowest. The coupling strength of several features was ordered as follows: tree > elevation > ground > sky > water > marker. The combination of features of the high-heat path is trees, height difference and ground. CONCLUSIONS: There is a significant correlation between the landscape features of park trail and the choice of trail. People prefer to choose the trail with good scenery, flat ground and wide view. The route selection trail with the application of the picture organization scheme has pleasant scenery and attraction, and can guarantee the training intensity, which provides a scientific basis for the design of the park trail landscape. Funding: This work was supported by the National Science Fundation of China [grant numbers 72104106]