| 摘要: |
| 城市人行道系统是城市各类活动的重要空间载体。科学评估人行道系统的步行环境,探索高效实时、具有可回溯性的评价方法对于提升城市出行品质、促进步行环境可持续发展具有重要意义。以人行道净化为视角构建一套人行道净化审计方法,制定人行道净化畅通性、安全性、舒适性三大类审计条目,运用无人机系统采集影像并构建高分辨率的二维正射影像和三维倾斜模型,通过虚拟审计平台构建审计数据库,以杭州市典型街区为例进行应用,分析人行道净化存在的问题和成因,形成改善清单、确定责任部门和估算整改资金。提出的评价方法具有数据收集时效性强、审计过程效率高、审计结果可复用的特点,以促进城市步行环境的健康、可持续更新建设。 |
| 关键词: 人行道净化 虚拟审计 无人机影像 倾斜模型 |
| DOI： |
| 分类号:TU984 |
| 基金项目: |
|
| Sidewalk Purification Audit Method Based on Unmanned Aerial Vehicle Imagery |
|
CHEN Mengwei,ZHUGE Bin,DENG Yiling,CHEN Siyi
|
|
Zhejiang University of Technology
|
| Abstract: |
| The urban pedestrian sidewalk system serves as a critical spatial carrier for diverse urban activities, directly influencing residents' walking experiences and the overall quality of urban travel. Scientific evaluation of the walking environment and the development of efficient, real-time, and retrospective auditing methods are essential for enhancing pedestrian infrastructure and promoting sustainable urban development. This study proposes a novel sidewalk purification audit method that integrates unmanned aerial vehicle (UAV) imagery, high-resolution two-dimensional (2D) orthophotos, and three-dimensional (3D) oblique models within a virtual auditing platform, addressing the limitations of traditional auditing approaches such as high labor costs, delayed problem reporting, and insufficient spatial resolution. The evaluation framework is structured around three core dimensions—sidewalk impediment, safety, and comfort—which are aligned with national guidelines and urban street design standards to address the fundamental needs of pedestrians. sidewalk impediment focuses on ensuring unobstructed pedestrian flow; safety involves identifying hazards, such as whether there is a barrier between the sidewalk and the bike lane, and whether there are safety hazards such as damaged facilities, uneven pavement, and missing or sunken manhole covers; and comfort considers the overall quality of the pedestrian experience, including the availability of resting and aesthetic features. To capture comprehensive urban data, UAV is employed to acquire detailed images across the study area. These images are then processed to generate accurate 2D orthophotos and 3D oblique models, which allow for precise measurement of key indicators such as effective sidewalk width, the presence of encroachments by motorized and non-motorized vehicles, and the condition of essential features like tactile paving. The integration of computer vision algorithms further enhances the process by automatically detecting and quantifying some issues, thereby improving both the efficiency and objectivity of the audit. A practical demonstration of the method is provided through a case study in Hangzhou. A 1-square-kilometer area surrounding a metro station—encompassing schools, commercial complexes, residential districts, and bus stops—was selected as the study area. Within this zone, the sidewalk network was segmented into uniform 20-meter sections to facilitate detailed, systematic evaluation. UAV flight routes were planned to ensure optimal image capture, even in locations where traditional street-level photography might be obstructed or outdated. This comprehensive dataset was then used to assess various indicators related to sidewalk impediments, safety, and comfort issues. The audit revealed several prevalent issues. In terms of encroachment, there is widespread e-bike occupancy, and many road segments lack designated parking spaces for these vehicles. Pedestrian safety is generally good, but the safety of passage for people with disabilities is often overlooked. In some road segments, continuous turns and unnecessary bends in tactile paving may confuse visually impaired pedestrians during their walk, while discontinuous tactile paving increases their safety risks. Excessive poles and boxes, combined with the lack of resting facilities or landscape features, result in a low level of pedestrian comfort. Based on the audit findings, the study proposes targeted improvement measures. Recommendations include reconfiguring non-motorized vehicle parking to free up sidewalk space, optimizing the design and placement of street furniture and tree pits to enhance effective widths, and standardizing the implementation of tactile paving to ensure both functionality and safety. The virtual auditing platform is designed to generate detailed improvement lists, assign responsibilities to relevant urban management departments, and even estimate the funds required for the proposed enhancements. These features enable strong operability in multi-department collaboration. Another advantage of the proposed method is the reusability of the collected digital data. The UAV images, orthophotos, and 3D models are stored in an audit database that supports ongoing monitoring and retrospective analysis. Urban planners and decision-makers can compare historical and current data to assess the impact of implemented improvements, promoting a continuous evaluation and enhancement cycle. Despite its strengths, the methodology has certain limitations. Some indicators of sidewalk conditions, such as effective sidewalk width and whether there are safety hazards such as damaged facilities, uneven pavement, and missing or sunken manhole covers, still require manual audit. Future research will aim to integrate more advanced computer vision techniques and detective models to reduce manual input further and enhance overall efficiency. In summary, this study presents a comprehensive and scalable sidewalk purification audit method that leverages UAV imagery and virtual auditing techniques to provide detailed, objective, and reproducible evaluations of urban pedestrian environments. The proposed evaluation method is characterized by the timeliness of data collection, high efficiency of the audit process, and reusability of the audit results to promote the healthy and sustainable renewal of the urban pedestrian environment. |
| Key words: Sidewalk purification Virtual Audit UAV Image Oblique Model |
