Pavement Distress Evaluation using 3D Depth Information from Stereo Vision

During the last few decades, many efforts have been made to produce automatic inspection systems to meet the specific requirements in assessing distress on road surfaces using video cameras and image processing algorithms. However, due to the noisy images from pavement surfaces, limited success was accomplished. One major issue with pure video based systems is their inability to discriminate dark areas not caused by pavement distress, such as, tire marks, oil spills, shadows, and recent fillings. To overcome the limitation of the conventional imaging based methods, a probabilistic relaxation technique based on 3-dimensional (3D) information is proposed in this report. The primary goal of this technique is to integrate conventional image processing techniques with stereovision technology to obtain an accurate topological structure of the road defects. In addition, a road scene often contains other objects such as grass, trees and buildings which should be separated from the pavement. Therefore the earlier algorithm has been enhanced to extract the pavement region from a road scene using a Support Vector Machine (SVM). Various types of cracks are then obtained from the pavement surface images and classified using a feed-forward neural network. The proposed algorithms are implemented in MATLAB and the results are presented. The second half of the document includes a report detailing the development of a software package that would allow the user to review digital photographs of pavement, evaluate that pavement by the PASER method, store the results in a database, and then make decisions based on the results of that analysis. The computer interface and analysis portion of the software was written in Microsoft Visual BASIC 2008. The long-term goals of this project include linking the evaluation results to a geographical information system database and developing various reporting strategies.