Perceptual Evaluation and Optimization of Streetscape Elements in the Ancient City of Suzhou,China:An Experimental Study based on Eye Tracking

LIU Changchun; LIU Mengfei; JIN Yumeng; WANG Dejing

doi:10.3969/j.issn.1000-0232.2025.12.011

Abstract：As a core component of urban public spaces,streets'environmental quality significantly affects the physical and mental health and well-being of residents,especially in ancient city streets supporting diverse activities and longer dwell times.Although the overall street style in Suzhou ancient city has been well-restored,there are still some problems with the streets'restoration,such as a lack of high-quality visual environments and insufficient comfort for residents.To shed further light on these issues,a case study based on Wusa Road and its branches in Suzhou ancient city was carried out.In the study,effects of streetscape elements on subjective and objective environmental perceptions of residents were analyzed systematically by eye tracking,virtual reality technology,and questionnaire-type surveys. First,streetscape images were collected using a panoramic camera.Plants,sky,buildings,roads,street furniture,and dynamic elements were accurately identified and quantified through semantic segmentation technology.Specifically,30 respondents were recruited to watch 20 representative street scenes in a controlled virtual reality environment.During the experiment,a helmet-mounted eye tracker was used to simultaneously collect respondents'average pupil diameter,cumulative fixation count at the Area of Interest(AOI),cumulative fixation duration at AOI,and cumulative visit count to AOI.Further,after they watched each scene,respondents'subjective perceptual evaluation ratings were collected across seven dimensions,including relaxation and stress relief. The results indicate that natural elements,especially plants,significantly enhanced subjective perceptual ratings across all dimensions(p<0.01).A moderate proportion of sky view is positively associated with comfort(p<0.05).Among artificial hardscape elements,high-density buildings trigger a sense of oppression,thus exerting a negative impact(p<0.05).However,moderate road width enhances the sense of safety and improves evaluation ratings for relaxation,stress relief,and so on(p<0.05).An increase in the amount of street furniture led to increased information processing load,thereby negatively affecting perceptual evaluations(p<0.05).Conversely,a moderate level of dynamic elements decreases feelings of fear or anxiety in unfamiliar environments(p<0.01).Scenes with a high proportion of plants,low building density,and few dynamic elements effectively reduce visual attention investment,whereas higher levels of sky visibility and wider roads trigger more visual exploration.Furthermore,the study found that a combination of high plant proportion;moderate proportions of sky visibility,road width,and street furniture;and low building density and dynamic-element proportion maximizes the environmental perception ratings of residents.Accordingly,Minzhi Road,with its high proportion of plants and open space,receives the highest subjective perception ratings,whereas sections of Wusa Road with dense buildings and mixed or high-volume pedestrian and vehicle flow receive the lowest ratings.Eye-tracking heat maps further indicate that scenes with rich building facades and displays of textual information attract more frequent and prolonged visual attention. Based on the above findings,the study proposes optimized design strategies for the street spaces in Suzhou ancient city.Renovations should prioritize the increase of green plants to shade excessive views of sky and buildings,integrate suggestive street furniture to reduce the visual cognitive load,and rationally control the traffic flow and pedestrian flow to improve residents'sense of safety.The research findings provide a scientifically informed theoretical basis for optimizing design of streets'visual environment in ancient cities.