Evaluation of the Visual Perception of Urban Single/Double-Layer Riverfront Greenway Landscapes Based on Deep Learning DOI Open Access
Xin Li, Yuan Wang, Zhenyu Wang

и другие.

Sustainability, Год журнала: 2024, Номер 16(23), С. 10391 - 10391

Опубликована: Ноя. 27, 2024

Urban inland rivers are closely related to urban development, but high-density urbanisation has reduced the natural function of streams and riverbanks hardened into two parts, embankment walls berms, which give rise a variety riparian landscapes. However, difference in height walkways affects degree their greening landscape effects. In this paper, we studied single- double-decker greenways, constructed quantitative indicators spatial elements based on deep learning algorithms using an image semantic segmentation (ISS) model that simulates human visual perception, used random forests multivariate linear regression models study impact riverfront greenway clarified differences caused by different types space aesthetic preferences (LP) confirmed specific extent components influence preferences. The results showed there were significant perception scores between single double layers. (1) WED (negative correlation) NI (positive is large single-layer greenway. colour, material structure guardrail can be beautified diversified quality greenery taken account maintain visibility order improve score (2) BVI double-layered positive. Water-friendly or water-viewing spaces added appropriately greenways. This applicable regional feature identification greenways large-scale hard barge bank images, realises whole-region perspective effective expansion analysis techniques for sustainable planning design

Язык: Английский

Exploring the Impact of Waterfront Street Environments on Human Perception DOI Creative Commons
Yiqing Yu, Gonghu Huang, Dong Sun

и другие.

Buildings, Год журнала: 2025, Номер 15(10), С. 1678 - 1678

Опубликована: Май 16, 2025

Urban waterfront streets are important mediators that reflect a city’s image and characteristics. They play positive role in enhancing residents’ cohesion, mental physical health, social interactions. Human perceptions represent individuals’ psychological experiences feelings toward the surrounding environment. Previous studies have explored impact of urban street-built environmental factors on perceptions; however, research focusing street environments their impacts human remains limited. Therefore, exploring specific characteristics different dimensions perception is essential for guiding development livable cities. Based Street View images (SVIs), this study applied artificial neural networks machine learning semantic segmentation techniques to obtain feature data Murasaki River line spaces Kitakyushu, Japan. In addition, correlation regression analyses were conducted explore features spaces, corresponding optimization strategies proposed. The results show greenness significantly enhances safety, wealth, beauty, while effectively reducing boredom depression. Furthermore, building visual ratio contributes increased vitality. On other hand, such as openness, spatial indicators, color diversity negative effects perceptions, including safety particular, openness increases This advances exploration from perspective perception, providing theoretical foundation improving quality offering references human-centered planning construction.

Язык: Английский

Процитировано

0

Evaluation of the Visual Perception of Urban Single/Double-Layer Riverfront Greenway Landscapes Based on Deep Learning DOI Open Access
Xin Li, Yuan Wang, Zhenyu Wang

и другие.

Sustainability, Год журнала: 2024, Номер 16(23), С. 10391 - 10391

Опубликована: Ноя. 27, 2024

Urban inland rivers are closely related to urban development, but high-density urbanisation has reduced the natural function of streams and riverbanks hardened into two parts, embankment walls berms, which give rise a variety riparian landscapes. However, difference in height walkways affects degree their greening landscape effects. In this paper, we studied single- double-decker greenways, constructed quantitative indicators spatial elements based on deep learning algorithms using an image semantic segmentation (ISS) model that simulates human visual perception, used random forests multivariate linear regression models study impact riverfront greenway clarified differences caused by different types space aesthetic preferences (LP) confirmed specific extent components influence preferences. The results showed there were significant perception scores between single double layers. (1) WED (negative correlation) NI (positive is large single-layer greenway. colour, material structure guardrail can be beautified diversified quality greenery taken account maintain visibility order improve score (2) BVI double-layered positive. Water-friendly or water-viewing spaces added appropriately greenways. This applicable regional feature identification greenways large-scale hard barge bank images, realises whole-region perspective effective expansion analysis techniques for sustainable planning design

Язык: Английский

Процитировано

0