- 408-924-7560
- mineta-institute@sjsu.edu
- Donate
Identify Pavement Cracking Using Machine Learning and Computer Vision
We have developed pavement condition survey using drone technology and a pothole photo submission and reporting program. We were able to obtain many images and videos. However, it is not very efficient using AutoCAD and manually identifying pavement distresses such as cracking and potholes. The computer vision and artificial intelligence have been greatly advanced over the past decade. Researchers and industry have been able to utilize these new technologies in image analysis, such as MRI, CT Scan, etc. We would like to use computer vision and machine learning to automatically identify pavement distresses including both type and quantity of pavement distresses. This automated approach promises to deliver greater accuracy, efficiency, and cost-effectiveness compared to traditional methods, revolutionizing pavement distress identification and supporting more proactive maintenance strategies.
Principal Investigator:
DingXin Cheng
PI Contact Information:
CSU Chico
Dates:
April 2024 to December 2024
Implementation of Research Outcomes:
We plan to develop a comprehensive prototype process that leverages computer vision and machine learning technologies to automatically identify, classify, and report potholes and pavement cracks.
This process will include the following key components:
- Automated Distress Detection System:
- A robust computer vision model capable of detecting various types of pavement distresses, including both surface-level cracks and deeper structural potholes.
- Integration of machine learning algorithms to classify distresses by type, severity, and extent.
- Quantification and Reporting Framework
- A system for accurately measuring the size and spread of detected distresses.
- Automated generation of detailed, geotagged reports that include distress types, locations, and severity metrics.
The research will contribute to a transformative approach to pavement distress management, enabling municipalities and transportation agencies to adopt more proactive and data-driven maintenance strategies.
Impacts/Benefits of Implementation:
The developed prototype has the potential to serve as a transformative foundation for enhancing pavement management systems, providing significant benefits to local agencies and the public, including:
- Improved Pavement Conditions
- Enabling more accurate and timely detection of pavement distresses, allowing for faster and more targeted repair actions.
- Reducing the backlog of untreated pavement issues, leading to improved road quality and a more reliable transportation network.
- Enhanced Public Safety
- Minimizing hazards associated with untreated cracks and potholes, thereby reducing the risk of vehicle damage, accidents, and injuries.
- Supporting proactive maintenance practices that prevent the escalation of minor distresses into major safety concerns.
- Cost Savings and Resource Efficiency
- Lowering maintenance costs through early detection and optimized repair strategies. Reducing reliance on labor-intensive and time-consuming manual surveys, freeing up resources for other critical tasks.
- Data-Driven Decision Making
- Providing local agencies with detailed, geotagged data on pavement conditions, enabling informed prioritization of maintenance projects. Facilitating long-term planning through the integration of distress trends and deterioration models.
- Environmental Benefits
- Reducing the environmental footprint of maintenance operations by optimizing repair schedules and minimizing unnecessary interventions.
- Contributing to sustainable infrastructure management through efficient resource utilization.
- Community Engagement and Satisfaction
- Demonstrating accountability and transparency through timely reporting and visible improvements in road conditions. Enhancing public confidence in local government initiatives aimed at maintaining safe and reliable roadways.
By implementing this prototype, local agencies can improve their pavement management systems that not only improves infrastructure quality but also ensures safer, more efficient travel for the public.
Project Number:
2435
-
Contact Us
SJSU Research Foundation 210 N. 4th Street, 4th Floor, San Jose, CA 95112 Phone: 408-924-7560 Email: mineta-institute@sjsu.edu
