AUTOMATIC DETECTION AND ANALYSIS OF SURFACE DISTRESSES USING A MULTI FUNCTIONAL VEHICLE

摘要

The automated detection of actual visual distresses such as different types of cracking, raveling etc is increasing fast in popularity due to the limitations and safety issues related to manual surveys. The fast developments in Line-Scan camera technology does result in high quality 2D images of the surface of a pavement. These developments have made it possible to analyse these images as well automatically with a high level of reliability. Recent developments to collect 3D images make it possible automatically detect and classify cracks, changes in texture and the related severity of raveling. The rating of the texture based on 3D images can be compared with the cumbersome manual Sand Patch Test, but with the 3D imaging, the total lane width and road section length is rated based on 250 x 250 mm squares. In addition segregation of a wearing course can be detected as well as edge drop-off. In this paper the system of the Multi Functional Vehicle (MFV) will be illustrated based on examples of 3D case studies. The combination with laser lighting makes it possible to survey pavement surfaces even at night without compromising the image quality. The rating of the severity and area of the distresses can be adjusted according to any existing visual distress procedure. Another main advantage is that all collected data is geo-referenced making it possible to either link the data to an existing GIS or export the results to Google Earth for optimum visualization of the location of each and any distress types.