Scholarlink

<p>Pedestrian cadence serves as an indicator for assessing urban pedestrian dynamics, reflecting not only the success of urban planning but also the rhythm of city life. This study employs computer vision and signal processing techniques to identify pedestrian cadence, with the aim of exploring its relationship with urban environments. The research methodology involves three primary components: Real-Time Image Collection, Pedestrian Feature Extraction, and Pedestrian Cadence Estimation. The Real-Time Image Collection Module autonomously gathers video streams from a network of public live cameras; the Pedestrian Feature Extraction Module employs the You Only Look Once version 7 (YOLOv7) model along with the Simple Online and Real-Time Tracking (SORT) algorithm to produce time series data of pedestrian characteristics; following this, the Pedestrian Cadence Estimation Module applies signal processing to these data to evaluate pedestrian cadence. This data collection and analysis approach is non-invasive, preserving pedestrians’ natural behavior, thereby enhancing data authenticity and reliability. This approach has been applied using various live public cameras to evaluate pedestrian cadence, examining differences between diverse geographical and functional zones (residential, educational, touristic, business, and traffic areas). Using K-Means clustering in pedestrian cadence cumulative distribution function (CDF) data from these regions, the study investigates the drivers behind these clusters. In addition, the research explored the connection between economic status and pedestrian cadence in different Taiwanese cities. Overall, this study provides urban planners and policymakers with valuable knowledge to promote smart city development and improve urban planning effectiveness and efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Urban pulse: measuring pace of life through public camera-based cadence analysis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月7日

查看全文
  • Pei-Yu Ho,
  • Ming-Kuang Chung,
  • Ling-Jyh Chen

摘要

Pedestrian cadence serves as an indicator for assessing urban pedestrian dynamics, reflecting not only the success of urban planning but also the rhythm of city life. This study employs computer vision and signal processing techniques to identify pedestrian cadence, with the aim of exploring its relationship with urban environments. The research methodology involves three primary components: Real-Time Image Collection, Pedestrian Feature Extraction, and Pedestrian Cadence Estimation. The Real-Time Image Collection Module autonomously gathers video streams from a network of public live cameras; the Pedestrian Feature Extraction Module employs the You Only Look Once version 7 (YOLOv7) model along with the Simple Online and Real-Time Tracking (SORT) algorithm to produce time series data of pedestrian characteristics; following this, the Pedestrian Cadence Estimation Module applies signal processing to these data to evaluate pedestrian cadence. This data collection and analysis approach is non-invasive, preserving pedestrians’ natural behavior, thereby enhancing data authenticity and reliability. This approach has been applied using various live public cameras to evaluate pedestrian cadence, examining differences between diverse geographical and functional zones (residential, educational, touristic, business, and traffic areas). Using K-Means clustering in pedestrian cadence cumulative distribution function (CDF) data from these regions, the study investigates the drivers behind these clusters. In addition, the research explored the connection between economic status and pedestrian cadence in different Taiwanese cities. Overall, this study provides urban planners and policymakers with valuable knowledge to promote smart city development and improve urban planning effectiveness and efficiency.