top of page

Integrating Non-Motorist Facility Data into Comprehensive Road Safety Assessment

Project Description

This project aims to enhance pedestrian and bicyclist safety understanding, bolster educational and professional capacities, and facilitate the practical implementation of computer vision techniques in transportation planning and engineering, ultimately leading to safer transportation systems for pedestrians and cyclists. The project will address challenges in gathering pedestrian facilities data and assessing safety concerns for pedestrians and bicyclists, encompassing a comprehensive process involving literature review, case study design, data collection and preparation, model development and validation, result analysis, and recommendation formulation.


Through innovative approaches utilizing satellite images and image processing techniques such as spatial analytics and deep learning models, the project intends to extract crucial information about pedestrian and bicyclist facilities and nighttime streetlight conditions. By leveraging deep neural networks and statistical analysis, the project aims to compare longitudinal datasets, predict injury risks, identify high-risk areas, and unravel potential risk factors and relationships contributing to pedestrian and bicycle accidents, thereby informing evidence-based decisions and interventions. The outcomes of this project will be disseminated through technical reports and academic discussion, contributing to the understanding of non-motorist safety, encouraging further research, and providing educational resources for transportation programs.


This project aims to develop software for retrieving satellite images online at an appropriate zoom level. The detailed image annotation procedure will be thoroughly explained. The team will create multiple learning models to fulfill the project's objectives. All software and algorithms will be developed using open-source tools, ensuring easy implementation, deployment, and transferability. Furthermore, the project will demonstrate practical applications of the collected data in enhancing pedestrian and bicycle safety. A comprehensive data inventory for pedestrian and bicyclist facilities will be developed as part of this initiative.


The final deliverable will be a comprehensive technical report that documents the data retrieval process. Moreover, the report will feature an in-depth statistical analysis, showcasing the correlation between pedestrian crash areas and these facilities. This analysis will include essential tables, graphs, and maps for clarity. Additionally, the team will provide valuable recommendations based on the insights and experiences gained throughout this project, aimed at continuously improving future research endeavors.


The project will document its findings and methodologies in a technical report. Specifically, the use of cutting-edge computer vision algorithms and satellite imagery to detect and evaluate non -motorist facilities. The project aims to disseminate these insights through academic channels to drive further advancements in this domain. Additionally, it is expected that the project will generate valuable insights into non-motorist safety, fostering ongoing research and innovative solutions, while also providing an educational resource for transportation programs.

The technical report, along with associated materials, can be seamlessly integrated into coursework, offering educators and students practical applications of computer vision and deep learning techniques. Furthermore, the project will provide a tool or software for non-motorist facility data retrieval, which will help transportation authorities, urban planners, and academic institutions in using related methodologies efficiently.


06/01/2023 to 05/31/2024


University of Wisconsin-Milwaukee

Principal Investigator

Xiao (Shaw) Qin

University of Wisconsin-Milwaukee

ORCID: 0000-0003-0073-3485

Research Project Funding

Federal: $65,074

Non-Federal: $32,503

Contract Number


Project Number


Research Priority

Promoting Safety

bottom of page