UAV Photogrammetry and OBIA Classification for Decay Mapping in Built Heritage

One of the first steps in the development of BIM (Building Information Model) for existing structure is linked to the acquisition and mapping of decay information. Currently, the most widely used practice is visual inspection carried out by experts in the field an based on visual inspection, that requires significant effort in terms of time, human resource and operational costs. To overcome many of the current challenges, the use of Unmanned Aerial Vehicle (UAV) photogrammetry combined with Object-Based Image Analysis (OBIA) techniques represents an innovative approach for non-invasive methodology for decay mapping. The paper presents the results of the integrated approach applied on two case study, differing in types of decay observed and materials. The first involves a building's façade mainly affected by crusts due to washout, and the second concerns a section of a concrete infrastructure, which is affected by different types of surface degradation (cracks, spalling, exposed bars and stains). The results were validated using different approaches in function of the decays observed: in the first case study a cross validation was performed and in the second one a confusion matrix from a ground truth was extracted. The results show satisfactory performance on the OBIA classification in both case study. The methodology developed overcomes some limitations of traditional methods and, the obtained maps, can be integrated in BIM models to enrich the informative content and subsequently support the development of strategies for improving the energy efficiency of the built environment.