Work

Can urban vitality be seen? Video analytics of social interaction and land use

Year: 2025

Type: article

Abstract: Social interaction is a powerful indicator of the vitality and quality of urban public spaces. In areas such as urban squares, the frequency, duration, and spatial distribution of interpersonal encounters reflect how effectively a space facilitates social interactions within the urban land system. However, these interactions—often brief and subtle—are difficult to observe and quantify in dynamic environments using conventional planning tools, limiting the proper management and design of urban spaces. To address this gap, this study proposes a human-centred, AI-based video analytic approach using computer vision techniques to detect and quantify social interactions. Interactions are classified into four categories—transient, sustained, long-term, and gathering—based on proximity and temporal thresholds, and are analysed in relation to spatial configurations, such as seating, greenery, and visual landmarks. A Social Interaction Intensity Index (SIII) is introduced to evaluate the capacity of different spatial zones to foster social activity, based on identified interaction points and their duration, normalized by area for fair comparison across zones. Rather than interpreting behaviour semantically, the index provides an objective and scalable indicator of how effectively urban land supports social functions and public life. This framework bridges data-driven sensing and human-centred land use policymaking, enabling planners and decision-makers to identify socially underperforming zones, guide spatial reconfiguration, and promote inclusive, evidence-based regeneration strategies that uphold humanistic values in public space design. • Computer vision enables rule-based automatic detection to quantify social interactions in public spaces. • A Social Interaction Intensity Index (SIII) compares vibrancy across different spatial zones. • Interaction types and their spatiotemporal patterns are linked to key urban features. • Provides a data-driven foundation for human-centred public space design and land use optimization.

Source: Land Use Policy

DOI: 10.1016/j.landusepol.2025.107818

Authors: Wenjie Chen, Asya Natapov, Yasir Ali

Institution: Loughborough University

Language: English

Topic: Urban Design and Spatial Analysis

Subfield: Building and Construction

Field: Engineering

Domain: Physical Sciences

SDG: Sustainable cities and communities

Open Access status: hybrid

Estimated APC paid: $3,830