Automated Real-Time Pavement Distress Analysis

Abstract

Pavement distress detection and analysis is an important component of pavement management systems. Conventional visual and manual pavement distress analysis techniques are very costly, time-consuming, dangerous, labor-intensive, tedious, and subjective; have a high degree of variability; are unable to provide meaningful quantitative information; and almost always lead to inconsistencies in distress detail over space and across evaluations. In this paper, a novel system for multipurpose automated real-time pavement distress analysis based on fuzzy logic and morphology is presented. The proposed system provides high data acquisition rates; effectively and accurately identifies the type, severity, and extent of surface distress; improves the safety and efficiency of data collection; offers an objective standard of analysis and classification of distress; helps identify cost-effective maintenance and repair plans; provides images and examples through the information highway to other user-researchers; and provides an image and sample bank for training or as the benchmark for testing new algorithms. The proposed system can also contribute to other research in pavement maintenance, repair, and rehabilitation.