Artificial Intelligence and Mobile Phone-Based Pavement Marking Condition Assessment and Litter Identification [Research Brief]

Details

• Creators:
  Chen, Jianli ; Kuang, Biao
• Corporate Creators:
  University of Utah. Department of Civil and Environmental Engineering
• Corporate Contributors:
  Center for Transformative Infrastructure Preservation and Sustainability (CTIPS) Region 8 University Transportation Center (UTC) ; United States. Department of Transportation. University Transportation Centers (UTC) Program ; United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology ; Utah Department of Transportation
• Subject/TRT Terms:
  Artificial Intelligence Detectors By Object Of Detection Litter Object Counting Object Detection Object Geolocation Risk Assessment Road Markings Roadside Hazards Roadside Litter And Trash
• Publication/ Report Number:
  CTIPS-25-002
• Resource Type:
  Brief
• Geographical Coverage:
  United States
• Edition:
  Research Brief
• Corporate Publisher:
  Center for Transformative Infrastructure Preservation and Sustainability (CTIPS) Region 8 University Transportation Center (UTC)
• Abstract:
  This research developed advanced AI-based methods for automated inspection of transportation assets, focusing on pavement markings and roadside litter. The methodology involved expanding and annotating two large-scale image datasets, each containing more than 6,000 self-collected images. Faded pavement markings were classified into white and yellow categories, while roadside litter was divided into four classes: white litter, black litter, leaves, and dirt. Two deep learning detection models were trained using the You Only Look Once (YOLO) architecture. In addition to object detection, researchers developed a counting algorithm to quantify the number of detected objects within roadway segments or video clips. A geolocation model was also created by integrating GPS data through time-based interpolation, achieving precise spatial localization of detected assets. Finally, an interactive visualization interface was implemented using the Folium Python library to display the georeferenced inspection results, enabling intuitive mapping and data-driven asset management.
• Format:
  PDF
• Funding:
  CTIPS-007
• Collection(s):
  University Transportation Centers
• Main Document Checksum:
  urn:sha-512:ed27a9db002e0879058fddf62b62192bf93a887d84cf2ee5acb0cd1ccb6a38e4f616d6c06c1d50a0e6889970e5cff341224f352beeb8e91fb8eb7945aced7c24
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/87797/dot_87797_DS1.pdf
• File Type:
  [PDF - 788.39 KB ]