Objective To clarify how urban visual environments influence human responses, this study reviews empirical evidence on urban visual landscape indicators and their effects. Urban visual environments play an important role in people-oriented urban development and healthy city planning. Street-view imagery, semantic segmentation, and computer vision now support large-scale measurement of human-scale visual exposure. Common indicators include green view index, visible sky ratio, visible building ratio, visible water ratio, enclosure, and openness. However, existing studies use different definitions, data sources, spatial scales, statistical models, and response variables. Their findings often differ in effect direction and strength. This limits comparison across studies and weakens their value for planning practice. To address this problem, this study constructs an “urban visual landscape－human response” analytical framework. It aims to identify the evidence distribution, effect direction, contextual differences, and vegetation-related mechanisms of urban visual landscape effects.

Methods This study followed the PRISMA framework. We searched the Web of Science Core Collection, Scopus, and CNKI databases for empirical studies published from January 2016 to January 2025. After screening titles, abstracts, and full texts, we included 119 studies. We standardized and coded visual indicators and human response variables. Urban visual landscape indicators were divided into four groups: color and lighting features, landscape elements, spatial form, and landscape imagery. Human responses were divided into six groups: restoration and stress, emotional and psychological states, safety and risk perception, aesthetic and preference evaluation, perceived comfort and environmental quality, and behavioral responses and use of space. To improve comparability, we unified effect directions. A positive effect means that a visual indicator relates to a more favorable human response. A negative effect means that it relates to a less favorable response. We used vote counting based on effect direction to summarize the overall pattern. We also calculated a net direction index for each visual indicator. Stratified analyses by landscape type and evidence quality tested the robustness and contextual boundaries of the findings. In addition, vegetation-related indicators had the largest evidence base. Therefore, we conducted a random-effects meta-analysis to compare visible vegetation ratio, species richness, structural diversity, and color diversity.

Results Results show that current research on urban visual environments is uneven. Existing studies mainly focus on landscape elements and spatial form. Color and lighting features and landscape imagery receive less attention. Among all indicators, visible vegetation index is the most frequently studied variable. It far exceeds visible sky ratio, visible building ratio, visible water ratio, enclosure, and openness. This pattern reflects the strong influence of street-view imagery and semantic segmentation methods. Different visual indicators show distinct relationships with human responses. visible vegetation index shows a stable positive association across response categories, landscape types, and evidence quality levels. Its net direction index is 0.73−0.74. Visible building ratio shows a relatively stable negative association. Its net direction index is −0.43. These two indicators act as relatively stable object-based visual cues. Vegetation usually conveys naturalness and restorative qualities. Buildings often convey artificiality, density, and visual pressure. Other indicators show stronger context dependence. Visible sky ratio, visible water ratio, enclosure, and openness do not show a single consistent direction. Their effects vary across landscape types and response categories. This indicates that their influence depends on spatial organization, scene type, and the specific human response being studied. The stratified analysis further shows the contextual boundaries of visual landscape effects. visible vegetation index relates positively to human responses in streets, residential communities, parks, urban green spaces, waterfront areas, and blue-green corridors. Visible building ratio generally relates to less favorable responses across contexts. In contrast, visible sky ratio may show positive effects in residential and waterfront settings, but negative or mixed effects in street spaces. Enclosure and openness also show mixed directions. These findings suggest that stable semantic cues and context-dependent spatial cues coexist in urban visual landscape effects. The vegetation-focused meta-analysis shows that vegetation effects should not be simplified as “more greenery is always better.” visible vegetation index represents the quantity of visual green exposure. It shows broad positive effects across several response dimensions, especially perceived comfort and environmental quality, aesthetic and preference evaluation, behavioral responses and use of space, restoration and stress reduction, and emotional and psychological states. Vegetation quality indicators show more specific effects. Structural diversity relates strongly to emotional restoration, with a pooled correlation coefficient of r = 0.715. Species richness relates closely to perceived environmental quality, with a pooled correlation coefficient of r = 0.646. Color diversity relates more closely to restoration experience and aesthetic evaluation. These results show a hierarchical relationship of “quantity exposure－quality experience－response differentiation”.

Conclusion This study provides a systematic synthesis of empirical evidence on urban visual landscapes and human responses. The findings indicate thaturban visual landscape effects are not fully linear or fully consistent. Stability and context dependence coexist. Visible vegetation and visible buildings are relatively stable visual cues. Sky visibility, water visibility, enclosure, and openness are more context-sensitive indicators. Based on these findings, we suggest that planning practice should avoid simple one-directional optimization of visual indicators. Urban greening strategies should not only increase visible vegetation coverage. They should also improve species richness, structural diversity, and color variation. Context-dependent indicators should be adjusted according to specific spatial settings and response targets. Overall, this study offers a reusable analytical framework for urban visual environment research and evidence-based implications for health-oriented urban planning and landscape design.