ABSTRACT
The condition evaluation of roadway transport networks is conducted to provide decision support for appropriate maintenance activities, preventing the possibility of detrimental effects. The costly, time-consuming and subjective current pavement assessment methods lead to the requirement for automation of the underlying process. Presented herein is an automated methodology for pavement patches detection; a process which is crucial for pavement surface evaluation and rating. Support Vector Machine (SVM) Classification is utilized, whilst the possibility of collecting pavement frames from smartphones, positioned insides of cars is examined. The SVM is trained and tested by feature vectors generated from the histogram and two texture descriptors of non-overlapped square blocks, which constitute an image. The outcome is the indication of the frames that include patches and the image blocks which are characterized as parts of patches.
