Artificial Intelligence (AI) Frameworks for Detecting Roadway Features Along Arterial Roadways from Planimetric Satellites Imagery Data

Departments of Transportation (DOTs) at the state level play a vital role in collecting and maintaining highway inventory data to support informed decision-making across various operational levels. Traditionally, these efforts have relied on labor-intensive and expensive processes, presenting challenges in updating and expanding inventory coverage. However, advancements in Artificial Intelligence (AI), particularly in computer vision and deep learning, offer a transformative solution to these limitations. The overall goal of this proposed research is to develop an AI-driven framework that enables automated extraction of roadway geometric features (i.e., pedestrian crosswalks and turn lanes) from aerial imagery, advancing Ohio Department of Transportation's (ODOT's) efficiency in data collection. The associated objectives include: (1) Collect and pre-process georeferenced aerial images for detecting pedestrian crosswalks and turn lanes, annotating them to train AI algorithms effectively. (2) Design and train a deep learning model to detect specific roadway features from satellite images. (3) Develop a Geographic Information System (GIS) database to organize and store the extracted features for easy accessibility and integration with existing ODOT datasets. (4) Build a flexible framework to support future expansion, enabling the detection of additional roadway features as needed.

• Record URL:
  • https://www.dot.state.oh.us/Divisions/Planning/SPR/Research/reportsandplans/Lists/Final%20Reports%20All/Item/displayifs.aspx?List=47f3581d%2Df21c%2D403b%2D9358%2Dfea0b008772b&ID=812&Source=https%3A%2F%2Fwww%2Edot%2Estate%2Eoh%2Eus%2FDivisions%2FPlanning%2FSPR%2FResearch%2Freportsandplans%2FLists%2FFinal%2520Reports%2520All%2FAllItems%2Easpx%3FuseFiltersInViewXml%3D1%26FilterField1%3DResearch%5Fx0020%5FOrganization%26FilterValue1%3DCleveland%2520State%2520University%26FilterType1%3DChoice%26FilterField2%3DList%5Fx0020%5FCategori%26FilterValue2%3DActive%26FilterType2%3DChoice&ContentTypeId=0x0100BD006C89430C884FB603A74E63BB6849&RootFolder=%2FDivisions%2FPlanning%2FSPR%2FResearch%2Freportsandplans%2FLists%2FFinal%20Reports%20All

Language

• English

Project

• Status: Active
• Funding: $150,000.00
• Contract Numbers:

  42150

  123341

  136986

• Sponsor Organizations:

  Federal Highway Administration

  1200 New Jersey Avenue, SE
  Washington, DC  United States  20590
• Managing Organizations:

  Ohio Department of Transportation

  Research Program
  1980 West Broad Street
  Columbus, OH  United States  43223
• Project Managers:

  Fout, Vicky

• Performing Organizations:

  Cleveland State University

  Euclid Avenue at 24th Street
  Cleveland, Oh  United States  44115
• Principal Investigators:

  Kidando, Emmanuel

• Start Date: 20250319
• Expected Completion Date: 20260919
• Actual Completion Date: 0

Subject/Index Terms

• TRT Terms: Aerial photography; Arterial highways; Artificial intelligence; Crosswalks; Data collection; Detection and identification systems; Geometric elements; Image analysis; Machine learning; Turning lanes
• Subject Areas: Data and Information Technology; Highways;

Filing Info

• Accession Number: 01948477
• Record Type: Research project
• Source Agency: Ohio Department of Transportation
• Contract Numbers: 42150, 123341, 136986
• Files: RIP, STATEDOT
• Created Date: Mar 14 2025 1:41PM