Image-based 3D reconstruction using traditional and mobile-phone data-sets for road pavement distress analysis

Abstract

The issue of road networks being in deplorable conditions is one that is widespread globally. One of the main precursors for this is that when preparing maintenance management systems, many road agencies rely on data which is often outdated or inaccurate. This is due in many cases to insufficient budgets which are unable to adequately address both maintenance and rehabilitation. It is therefore critical that road agencies have better tools at their disposal to help combat these issues. One of the possible techniques that have been identified is the use of structure from motion techniques to adequately identify road pavement distresses. This paper advances previous work in this area and explores the accuracy of using mobile phones to collect the imagery as opposed to traditional methods relying on professional cameras and equipment. This would provide a lower cost and readily available alternative for practitioners. The techniques have been applied on a distressed pavement in Palermo, Italy using data-sets from a mobile phone and a professional camera to analyse the quality and adequacy of using data-sets from the mobile phone. The results indicate that the mobile phone data-sets can adequately utilize the techniques and therefore this incites the possibility of integrating mobile integrations with the technology specifically focused on pavement management systems.