Abstract

Pavement condition monitoring is essential for pavement maintenance programs. Current practice includes the usage of specialized vehicles for collecting pavement data. However, the high purchase and operational costs restrict their application in the primary road network. Therefore, manual visual surveys are extensively applied as well. Moreover, raw data is mostly analysed manually adding further to labour intensiveness and time consumption of the whole process. In order to overcome the aforementioned limitations, the focus of this research is to propose a method that will automatically detect and classify pavement defects. The general idea is to treat the problem of pavement assessment holistically, use data from sensors that either pre-exist on vehicles or are easy to purchase and attach, in order to transform passenger vehicles into ubiquitous pavement monitoring ‘reporters’. Initially, a method that detects and tracks pavement patches in image and video data captured from a parking camera was created. The results show that the use of such data is quite promising. The overall proposed approach will provide a way of reporting road conditions inexpensively, globally and frequently.