Description:
- Pavement condition assessment
Abstract
USC researchers have developed a cost-effective solution for autonomously assessing road-surface condition that utilizes an RGB color image and an infrared projector and camera for depth sensing. The system can autonomously detect and quantify various road conditions such as patching, cracks, and potholes, and it employs a novel approach that does not require training. Field experiments show it can accurately locate detected defects by incorporating GPS information. This solution can serve as a supplementary sensor system in pavement surface assessment vehicles, reducing operational costs.
Benefit
- Autonomous detection and quantification of road surface condition
- Utilizes inexpensive RGB-D sensor system
- Sensor data interpretation does not require training
- Accurately localizes detected defects
Market Application
Pavement assessment procedures in the United States are time- consuming, labor-intensive, expensive, and can lead to safety threats for personnel involved. However, many automated approaches lack consistency and struggle with non-crack patterns, while expensive laser scanning systems have limited adoption. Automated defect assessment utilizing image processing and depth information technologies can address these challenges, providing accurate and reliable pavement surface assessment to promote maintenance and reduce cost.
Publications
Unsupervised Approach for Autonomous Pavement-Defect Detection and Quantification Using an Inexpensive Depth Sensor, Jahanshahi et al., 2012.
Other
Stage of Development
- Experimentally validated in field tests
- Available for licensing