Multi-Resolution Information Mining and a Computer Vision Approach to Pavement Condition Distresses


Download Final Report

CAIT project no.: CAIT-UTC-009

Fiscal Year: 2012/2013

Status: Final

Rutgers-CAIT Author(s): Nii Attoh-Okine, Ph.D., P.E., University of Delaware , Patrick Szary, Ph.D.

External Author(s): Dr. Alexander Appea, P.E

Sponsor(s): Virginia Department of Transportation, FHWA - RITA

Summary:

This research outlines three primary goals. First, develop a robust vision system addressing three key challenges of traditional pavement distress detection systems. Second, perform the integration of the vision system on to a GIS platform for crack classification and quantification. GIS will be used to generate real time road condition maps and provide recommendations regarding maintenance actions. Third, establish a real-time implementation of the system through parallel processing on current generation of multi-core CPUs.