Morphological Analysis in Infrastructure Engineering

by Chakravarthy Bhagvati, Rensselaer Polytechnic Inst, Troy, United States,
Dimitri A. Grivas, Rensselaer Polytechnic Inst, Troy, United States,
Michael M. Skolnick, Rensselaer Polytechnic Inst, Troy, United States,

Document Type: Proceeding Paper

Part of: Digital Image Processing: Techniques and Applications in Civil Engineering


This paper presents applications of morphological image analysis techniques in infrastructure engineering. Emphasis is placed on pavement surface condition analysis and the classification of various categories of pavement distress in terms of severity and extent. It examines the ability of morphological size distributions to distinguish variations in texture indicative of distress from inhomogeneities in texture itself. Current work investigates texture models based on the distribution of area and number of particles at different scales where the distributions are assumed to be gaussian. Preliminary findings indicate that the developed method is superior to existing approaches in its ability to handle several types of distresses (like raveling, surface wear, etc.) in addition to cracking. Results are presented on images of asphalt and Portland cement concrete surfaces that have been previously classified by human experts, and important discussions on the performance of the developed techniques are presented and discussed.

Subject Headings: Concrete pavements | Asphalt pavements | Surface properties | Infrastructure | Computer vision and image processing | Pavements | Pavement condition

Services: Buy this book/Buy this article


Return to search