Robust Method for Automated Segmentation of Frames with/without Distress from Road Surface Video Clips

Abstract

Automated processing of road surface video clips captured for road condition assessment is necessary to detect the existence of road surface distress in less time and efforts. This paper presents a robust method for automated segmentation of frames with/without distress from road surface video clips captured by existing camera based imaging systems without any artificial lighting systems. The proposed method is based on an adaptive thresholding technique and user defined decision logic for automated detection of road surface distresses out of such video clips. This method has been implemented in a Windows Vista environment with the help of Visual Studio 2008 and OpenCV library and tested on 31 road surface video clips of Indian Highways. The results show that this method can segment a road surface video clip automatically into two categories of video frames, namely frames with distress and frames without distress, with accuracy up to 96% while saving a considerable amount of time and manpower resources.