Objective Street space, as an indispensable landscape space, is deeply embedded in the urban texture and has become an important carrier for citizens’ social life. It is not only a channel for traffic flow, but also a source of urban vitality and a place for citizens to place their emotions. Throughout history, street space design theory has gone through the solemnity of classicism, the pursuit of efficiency of functionalism, and the exploration of harmonious coexistence between ecology and nature, and has, under the guidance of humanistic thought, given birth to design concepts such as “vibrant street” and “shared street” with citizen experience as the core. These concepts have profoundly reshaped our cognition and expectations of street space. The research on crowd activities in street space aims to reveal the types and characteristics of activities, understand the inherent connection between design and behavior, and provide a scientific basis for street environment optimization and resource allocation. This will not only help activate urban vitality, but also lead street space to develop in a more humane and diversified direction. In order to accurately match the urban space landscape design with the actual activity needs of diversified crowds, this research aims to overcome the limitations of existing digital technologies in large-scale macro-activity analysis, deeply explore the crowd behavior patterns at the landscape scale, and provide new perspectives and tools for understanding the complex interactive mechanism between the built street environment and the differentiation of crowd activities, thereby guiding the optimization and upgrading of urban street space.

Methods With the rapid development of artificial intelligence algorithms and the location-based services (LBS) data technology, this research makes full use of these advanced technologies and conducts an empirical study in a typical area of Zhongyang Road along the Central Axis of Nanjing. By collecting and analyzing fine spatiotemporal data based on mobile phone LBS and street space data, the research uses advanced artificial intelligence algorithms to deeply explore the activity characteristics of different crowds in street space. This method not only covers basic attribute analysis, but also deeply explores multi-dimensional indicators such as time preference, space preference, facility preference and landscape preference of crowds, laying a solid foundation for the measurement and genealogy of digital portraits of crowds in street space.

Results This research successfully constructs a digital portrait system of crowds in street space containing multi-dimensional indicators, and reveals 213 different types of crowds in street space and their behavioral characteristics in Zhongyang Road along the Central Axis of Nanjing through data analysis. Furthermore, the research summarizes the digital portraits of 7 typical crowds, including selecting portrait groups with a size of more than 2% of the total population from 213 types of digital portraits as typical groups of street space users. The specific portrait categories include the MOBD crowd featuring long duration, morning, open space, shopping and entertainment, and no preference, the ITCD crowd featuring medium duration, evening, compact space, catering service, and no preference, the NOBD crowd featuring long duration, noon, open space, shopping and entertainment, and no preference, the AGCD crowd featuring long duration, afternoon, general space, catering service, and no preference, the NOWD crowd featuring long duration, noon, open space, office service, and no preference, the MOBS crowd featuring long duration, morning, open space, shopping and entertainment, and shade preference, and the NOBP crowd featuring long duration, noon, open space, shopping and entertainment, and plaza preference. These portraits not only reflect the spatiotemporal use patterns of crowds in street space, but also reveal the underlying differentiation characteristics. In addition, this research also deeply analyzes the correlation between these crowd portraits and the current street landscape design and facility layout, providing strong support for understanding the interaction between street space environment and crowd activities. Further, this research takes Zifeng Block as an example to analyze the aggregation and flow characteristics of crowds in street space. Through portrait clustering and feature analysis, the intrinsic connection between the tempospatial differentiation of the crowds and factors such as individual attributes, built environment and time is revealed. The research finds that the crowds in the street space of Zifeng Block mainly gather near the subway entrances and exits or branch stations, and there exist significant differences in the composition of the crowds in different time periods. On weekday mornings, the area dominated by commercial office buildings typically attracts a large number of office workers, forming a long-term office service crowd; while in the evening and at night, the rich entertainment facilities attract more consumers. In addition, the crowds have a high demand for the quality of landscape space, but the accessibility of the existing square green space is limited, which affects their willingness to stay. The research also points out that Zifeng Block needs to further optimize the settings of street furniture, green landscape, etc., to improve the quality of street space and enhance its attractiveness to pedestrians. These insights are of great reference value for urban planning and decision-making, and can help create a more livable and business-friendly urban environment.

Conclusion This research not only enriches the research content on crowd behavior patterns in urban space, but also proposes a method for constructing digital portraits of crowds in street space based on fine spatiotemporal behavior measurement. By revealing the spatiotemporal differentiation characteristics of the digital portraits of crowds in street space and their correlation with the built street environment, this research provides a scientific basis and practical guidance for urban street design. Based on these findings, the research puts forward targeted planning response suggestions, aiming to guide the improvement of spatial quality by optimizing street space design, so as to better meet the actual activity needs of different crowds and promote the harmonious and sustainable development of the city.