Identifying the integrated visual characteristics of greenway landscape: A focus on human perception

Understanding the integrated landscape characteristics and spatial sequences of greenways and the surrounding visual environments is crucial for connecting fragmented natural spaces within cities and enhancing residents’ natural experiences. However, previous studies have mostly predominantly focused on individual perceptions of greenway landscapes, rather than considering the integrated visual characteristics of both the inner and the visible surroundings. This disparity often deviates from people's actual perception of these spaces. To address this gap, we have adopted a new approach to identify the integrated greenway visual characteristics in inner and surrounding landscapes from a human perception perspective, in the city of Wuhan, China. Our methodology involved the utilization of deep learning systems, text semantic analysis, and other techniques to integrate multi-source data, including street view images, remote sensing images, and perceptual comments, to quantitatively assess the landscape sequences and naturalness of landscape changes along the greenway paths. We identify the positions and characteristics of unique landscape nodes within the greenway. The findings of our study revealed ten main landscape types of greenways in Wuhan, with over 72 % of the greenways comprising more than five types. As the length of the greenways increases, the number of landscape types initially experiences rapid growth before reaching a stable state. Furthermore, the frequency of landscape change exhibits a linear and positive correlation with greenway length. Notably, the naturalness of the majority of greenway inner landscapes closely aligns with that of the surrounding visible environment. The openness of all greenways was positively correlated with the proportion of farmland and grassland in their surrounding areas. Furthermore, people exhibit higher positive sentiment responses to greenways that maintain a high level of natural consistency between the inside and outside environments. Additionally, over 47 % of the unique nodes primarily coincide with a specific landscape type and play a vital role in shaping the rhythm of greenway landscape change. These findings provide valuable insights to support the construction and management of greenway landscapes, as well as guidance for urban planners and designers aiming to harness natural resources effectively in the development of urban greenways.