الفهرس | Only 14 pages are availabe for public view |
Abstract Detection of road surface damages such as cracks and defects is very important. It helps to identify the required maintenance to prevent accidents, evaluate road maintenance budgets,and predict the remaining service lifecycle of a road. This thesis proposes a system for automatic analysis of road flexible pavement surface images structured into four stages, to address the problems of crack detection and crack typecharacterization, the system being able to characterize different crack types in a given image. Images are converted to gray for detection, then filtering to account for their suitability to reduce the high variance of pixel intensities found in typical road pavement surface images, without deteriorating crack information ,finally get threshold and characterize the crack into vertical (Region I), alligator or block (Region II), miscellaneous (Region III), and transverse crack (Region IV). Also modern survey equipment,the 3D laser scanner is utilized in the current study for defects detection and classificationand estimate the quantities required for maintaining. The developed software and the 3D laser scanner show high efficiency to detect and classify the road surface damages. |