Transportation Informatics: Advanced Image Processing Techniques for Automated Pavement Distress Evaluation

Principal Investigators

Dr. Ezzatollah Salari
The University of Toledo

Co-Principal Investigators

Dr James Lynch
University of Detroit Mercy

Co-Investigators

Dr. Eddie Chou
The University of Toledo
Dr. Utpal Dutta
University of Detroit Mercy

Project Dates:

01/01/2009 to 04/30/2010

Project year: Year 2

MIOH-UTC Project Identifier: TS 18; Project 2

Focus Area:

Research: Intelligent Transportation Systems

Abstract:

A multi-modal transportation system is a major asset to a region's economic vitality. Efficiency and safety of freight and passenger movements depends upon the accessibility of information such as available capacity and the physical conditions of the transportation infrastructures. Innovative information technologies can make gathering, processing, and exchanging of such information more effective, therefore, reducing unnecessary delays due to accidents or break-downs. As a result, overall transportation costs may be significantly reduced, making the area an even more competitive regional transportation hub, and helping to revitalize the regional economy.

Video technologies and image processing techniques used as a tool for inspection, monitoring, and diagnosis have long been adopted by various fields, most notably medicine and remote sensing. However, its use in transportation is still not widespread. The reason is that the amount of data to be processed could be very large and repetitive, while the accuracy requirements may not be as strict as in, say, medical diagnosis. Therefore, automated analysis and pattern recognition is highly desirable.

The development of an automated pavement inspection system involves two steps: a) the distress data acquisition, and b) the distress data interpretation. The major task of this project concentrates on the development of image processing techniques for surface anomaly detection and characterization. This process includes the recognition of abnormalities in an image and the extraction and quantification of the surface cracks. The focus will be on the analysis of pavement image data that will lead to the extraction of information from the images for transportation facility inspection, monitoring, and diagnosis purpose. Specifically, robust image processing and pattern recognition techniques will be devised to extract surface anomalies such as transverse, longitudinal, and block cracking for transportation applications. A typical distress inspection task can be divided into three stages: preprocessing, segmentation, and classification.