AN ASSESSMENT OF THE FEASIBILITY OF DEVELOPING AND IMPLEMENTING AN AUTOMATED PAVEMENT DISTRESS SURVEY SYSTEM INCORPORATING DIGITAL IMAGE PROCESSING

The rational allocation of pavement maintenance resources requires the periodic assessment of the condition of all pavements. Traditional manual pavement distress surveys, which are based on visual inspection, are labor intensive, slow, and expensive, and they pose a safety hazard when the raters have to get out of their vehicles and inspect the road on foot. One of the principal goals of this report is to provide the Virginia Department of Transportation with some of the important background required for determining whether, with the current state of development of image processing and camera and computer hardware, it is feasible to develop an automated system to use for pavement distress surveys. This report describes some of the fundamental techniques of image processing that are likely to play a role in a pavement survey system that can automatically recognize and classify cracks. The report concludes that developments in technology during the last 10 to 15 years have made it possible to develop and implement such a system and that the implementation of such a system would mean that surveys would be less expensive, faster, safer, and more objective.