1. Introduction
Historic districts represent a crucial aspect of urban cultural heritage, and their revitalization is essential for the sustainable development of cities [
1]. As one of the most vibrant micro-public spaces within historic districts, street corners differ from linear street spaces by simultaneously guiding pedestrian flow and gathering crowds [
2]. They serve not only as traffic nodes but also as vital spaces for social interaction and cultural expression [
3]. Therefore, optimizing street corner spaces has become a crucial strategy for revitalizing historic districts and enhancing their quality in the era of urban regeneration [
4]. An in-depth analysis of the factors influencing street corner vitality in historic districts holds significant value for improving both urban space quality and residents’ quality of life.
The relationship between the street environment and the vitality of the crowd has always been a hot issue in urban research. Jacobs believed that the street is the core of the modern city [
5] and took the lead in proposing the concept of street vitality. Lynch considered vitality a primary indicator for assessing urban spatial quality [
6], while Gehl emphasized that enhancing street vitality is key to improving urban life quality [
7]. Urban morphology experts view street vitality as a manifestation of activity stemming from the spatial structure of streets, asserting that the physical environment of buildings and streets shapes and influences street-level life and activity [
8]. The current research on the impact of the built environment on street activities is categorized into qualitative and quantitative studies. Qualitative methods include field surveys, questionnaires, and cognitive mapping [
9], such as Jalaladdini et al., who analyzed street vitality and its influencing factors using a questionnaire survey [
10]. However, qualitative methods are labor-intensive, limited in scale, and often lack timeliness, thus generally serving as a supplement to quantitative approaches.
Quantitative methods dominate the current research. With advancements in computer technology, methods integrating big data and computer vision are increasingly employed in street environment research. The quantitative analysis of the physical environment, incorporating street view images, POI data, and road network information from multi-source datasets, has become a standard approach [
11]; additionally, deep learning algorithms for semantic segmentation of street view images and target detection have advanced beyond traditional qualitative methods, enabling the quantification of street environment elements from a human perspective [
12], e.g., Hyung-Sup Shin et al. used a semantic segmentation algorithm to process a street view image dataset to anticipate environments suitable for crowd activities [
13].
Big data technology is also widely used to measure crowd vitality in urban environments, and the more common methods for obtaining crowd vitality data include population heat maps [
14], mobile signal data [
15,
16], GPS data [
17], LBS data [
18], Wi-Fi hotspot data [
19], traffic trajectory data [
20,
21], nighttime light data [
22,
23], and social media data [
24]. The above methods can achieve the purpose of obtaining crowd vitality data on a large scale, but they are not precise enough for small-scale studies and singly measure crowd vitality.
Furthermore, mathematical and statistical methods, particularly regression analysis, have been widely adopted in recent years to study the interaction between street vitality and the built environment, valued for their objectivity, clarity, and detail. For example, Sung constructed a multivariate regression model to explore the correlation between pedestrian activities and elements of the built environment in the case of Seoul’s streets [
25].
The aforementioned quantitative methods objectively and accurately measure street space attributes and examine the relationship between urban street crowd vitality and its influencing factors from a social perception perspective. In summary, there are abundant research results on the relationship between street environment and crowd vitality; however, the existing studies still have some shortcomings. The current body of research on urban environments has predominantly focused on large- and medium-scale analyses, such as neighborhoods or entire streets, often overlooking the nuanced dynamics of street corner spaces. This oversight is particularly evident in the dearth of studies dedicated to the quantitative analysis of environmental elements and crowd vitality within these micro-scale public spaces. Furthermore, while deep learning algorithms have been instrumental in processing street images for broader urban studies, their application to the quantitative examination of street corner spatial characteristics remains largely unexplored. Similarly, quantitative methods that have been effective in larger contexts, such as big data analytics, have proven inadequate for assessing crowd vitality in small-scale public spaces due to inherent limitations in data precision. Therefore, there is a clear need for the development of tailored quantitative research methods that can accurately capture and analyze the spatial elements and crowd dynamics unique to street corners.
Therefore, this study aims to examine the correlation between spatial environmental elements and crowd vitality on the street corners of historic districts. This study quantitatively analyzes the key spatial elements influencing street corner vitality using multi-source data acquisition and deep learning algorithms, employing a multidimensional, high-precision approach to measure crowd vitality in these spaces, and explores the interaction between spatial elements and crowd vitality in historic district street corners through multiple linear regression models.
5. Discussion
Based on the data results of the study, as a key element of corner spaces in historic districts, the form and function of corner buildings significantly influence crowd vitality (44.02%). In particular, the tourist nature of the Bund Historic District in Shanghai amplifies the impact of the historical quality of corner buildings on crowd vitality (18.57%). During the field research, it was also found that stopping in front of the historic buildings, filming, and talking were the main types of lingering activities. This finding aligns with previous studies [
29]. To enhance corner vitality, the preservation measures for historic buildings should be strengthened to ensure their historical qualities are maintained [
27]. Additionally, when the public accessibility and functional diversity of a corner building’s first floor are higher, crowd frequency increases, thereby enhancing the vitality of the corner space [
30]. The public accessibility of the first floor should be enhanced by adding functions such as cafés, bookstores, and exhibition spaces to further engage the public and promote crowd vitality in corner spaces [
41].
The form of corner spaces accounts for 37.13% of the influence on crowd vitality, with interface transparency being the most influential factor (13.96%). Corner spaces with highly transparent building interfaces tend to attract crowds, encouraging them to linger and engage in activities. Using transparent or semi-transparent materials to improve building interface transparency can increase the visibility and interactivity of the corner space, attracting pedestrian attention and prolonging their stay [
36]. Additionally, more open corner spaces (12.48% impact) and those with higher green visibility (10.69% impact) are associated with increased crowd activity. These two indicators influence spatial vitality by affecting visual perception and psychological states. Increasing the openness of corner spaces and introducing more green elements, such as trees, lawns, and vertical greening, can significantly enhance corner vitality [
12].
The functional influence of corner spaces on crowd vitality is 18.85%, with the density of functional facilities accounting for 11.16%. Functional facilities (e.g., street furniture, landscape elements) in corner spaces impact crowd activities by enhancing the space’s convenience and capacity for longer stays [
2]. To enhance corner vitality, additional street furniture, such as chairs and art installations, should be incorporated, providing spaces for rest and socialization. Additional facilities, such as vending machines and signage, can further enhance the convenience and attractiveness of corner spaces [
42]. Walkability (7.69%) primarily reflects pedestrian access to corner spaces and is the least influential of the eight factors [
12].
At the same time, there are limitations in the current study. Filming and field surveys of street corner spaces are conducted manually, limiting continuous data collection over extended periods. However, the research includes a one-week sampling period covering both weekdays and weekends, ensuring significant and representative temporal coverage. Processing street corner images with deep learning algorithms offers advantages, such as speed and objectivity, but also presents limitations. The limitations include information loss and the restricted applicability of certain indicators. Not all indicators are identifiable, definable, or calculable through deep learning algorithms. For instance, because of the accuracy limitations of semantic segmentation algorithms, transparency indicators must be obtained through manual measurement of basic data. Additionally, target detection faces challenges in accurately identifying the age characteristics of people in street corner photographs.
Consequently, future research should focus on increasing time efficiency, refining data collection methods, and broadening the scope of spatial element indicators. Additionally, subsequent studies should not only build upon the current framework but also evaluate its applicability in various urban environments and augment the dataset to improve the model’s predictive accuracy.
6. Conclusions
In urban conservation and revitalization, this study addresses a pivotal yet often overlooked aspect: the micro-scale dynamics of street corner spaces in historic districts. This research sheds light on the subtle interplay of factors contributing to the vitality of these spaces, thereby addressing a gap in the current scholarly discourse. An integrated approach was employed, utilizing various data sources to construct a nuanced model for assessing street corner vitality. The model is underpinned by field research capturing the essence of on-ground interactions, complemented by an expansive repository of web-based open-source data. Additionally, computerized deep-learning analyses were applied to quantify and interpret the spatial characteristics of street corners. Through multiple linear regression, the analysis identified the spatial elements exerting significant influence on street corner vitality. The regression model revealed the presence and magnitude of each element’s effect on crowd vitality.
It was found that corner building historicity, first-floor functional communality, transparency, openness, density of functional facilities, greenness, functional variety of building, and walkability have a significant impact on the vitality of the crowd in the corner space of the historic district, which can explain 77.5% of the vitality of the crowd in the corner space of the historic district.
The findings enrich quantitative research on micro-street environments and provide actionable insights. They shed light on the multifaceted nature of street corner spaces and emphasize their critical role in energizing historic neighborhoods. Thus, this research contributes a more comprehensive understanding of how small urban elements can significantly impact larger urban systems. Additionally, due to the generalizability and transferability of the indicators used in this study, the workflow remains applicable to similar research topics and holds value for reference and replication.
In essence, this study serves as a call to action for urban planners, conservationists, and researchers alike. It highlights the transformative potential of street corner spaces and encourages a reevaluation of strategies for preserving and enhancing the vitality of historic districts.