THE DIAGNOSIS OF ROAD SURFACE DISTRESSES THROUGH IMAGE-BASED MODELING TECHNIQUES. EXPERIMENTAL SURVEY ON LABORATORY-RUTTED SAMPLES

Abstract

In the past few years, there has been a drastic increase in the use of image-based modeling (IBM) techniques to create high quality, reality-based 3D models. The low costs of these techniques, as well as their attractive visual quality, have led many researchers and professionals to invest their energy and resources in several tests. IBM is rarely used in the field of road surface distresses as diagnosis is usually performed using other techniques and devices. Road safety statistics reveal that about a half of the total number of accidents occur mainly due to the deterioration of the pavement. The goal of effective road network management is often incompatible with economic resources designated for maintenance and rehabilitation. For this reason, IBM diagnosis of distresses seems necessary in order to both increase the level of road safety and to avoid incorrect interventions and treatments of road pavement. One of the strengthens of multi-view stereo techniques is the possibility to capture millions of points in a very short time, and to produce a 3D, textured polygonal model that can easily be used for visualizing and communicating digital assets. Our goal was to implement the IBM techniques on a laboratory-rutted sample and to verify the metric accuracy of the model and its validity for the distress diagnosis in terms of severity (rut depth). In order to assess the IBM technique, we compared its 3D model to the blue LED 3D scan (Artec Spider) of the same rutted sample.