Estimating Pavement Roughness by Fusing Color and Depth Data Obtained from an Inexpensive RGB-D Sensor.

Area of Science:

• Civil Engineering
• Geospatial Technology
• Sensor Technology

Background:

• Pavement management relies heavily on accurate roughness and defect detection.
• Current roughness measurement methods are often costly and labor-intensive.
• There is a need for affordable and precise road condition assessment tools.

Purpose of the Study:

• To develop and validate a cost-effective method for estimating pavement roughness using an RGB-D sensor.
• To propose an algorithm for autonomous quantification of road roughness from RGB-D data.
• To assess the accuracy and repeatability of the proposed sensor-based approach.

Main Methods:

• Calibration of an RGB-D sensor for outdoor use.
• Collection of synchronized RGB and depth data.
• Registration of RGB and depth frames to reconstruct 3D road surfaces.
• Development of an algorithm to process 3D road data for roughness estimation.

Main Results:

• A significant correlation was found between the International Roughness Index (IRI) estimated by the RGB-D sensor and manual measurements.
• Power Spectral Density (PSD) analysis confirmed the method's ability to capture road surface variations.
• Repeatability tests demonstrated the reliability of the proposed measurement technique.

Conclusions:

• The proposed RGB-D sensor-based system offers a cost-effective and accurate solution for pavement roughness estimation.
• The developed algorithm effectively quantifies road roughness, providing valuable data for pavement management.
• This technology presents a viable alternative to traditional, more expensive methods for road condition assessment.