Continuous Road Surface Distress Detection

Introduction State and local maintenance departments are tasked with keeping roads in good repair. One part of this task is to monitor the degradation of road surfaces, which manifests itself with the presence of cracks, potholes, and other distress. Currently, this is done by (Figure 1): (i) inspectors who visually judge the road condition, (ii) specialized vehicles which measure the distress with cameras or laser devices, or (iii) citizens who call in their observations. The first method is tedious and often inconsistent if several inspectors do the inspections. The second method does not have these problems, but it is generally expensive. The cost can range between $40 and $230 per mile. Because of budget constraints one stretch of road is usually only traversed by the specialized vehicle once every two years. This method is effective for interstates and highways, but it is not practical for the inner roads in a city. As for the third method, reports by citizens are generally only about severe problems, like large potholes on main roads. Each of the three methods has its strength, but none of them is able to monitor the road on a continuous basis at low cost with minimal human intervention.

[1]  Jitendra Malik,et al.  Contour and Texture Analysis for Image Segmentation , 2001, International Journal of Computer Vision.