Automated Real-Time Pavement Crack Detection and Classification System

摘要

Statistics published by the Federal Highway Administration indicate that the maintenance and rehabilitation of highway pavement in the United States requires over 17 billion dollars per year. Currently, maintenance and repairs account for nearly one-third of all federal, state and local government road expenditures. In the present manual systems, road crews walk along a given road with a vehicle following behind. When the crew finds cracks, they stop the vehicle and measure the cracks. While pavement monitoring and evaluation are essential requirements for effective pavement management, the manual systems described above are slow, costly, unsafe and subjective. Ideally, an automated crack detection system should detect all types of cracking and other surface distress of all sizes and at any collection speed. It should be affordable, easy to operate, and capable of daylight operation. Pavement management systems should provide meaningful, repeatable distress ratings to sections of pavement, thus supplying critical information for maintenance-related decision making. Much effort has been paid to the development of automated pavement crack detection algorithms and systems.