Digital image processing for automatic detection of cracks in buildings coatings

Abstract In the phase of diagnosis of pathological manifestations in facades, the visual inspection stage deserves special attention due to its inherent complexity (height, size, access difficulties and exposure conditions). In recent years, there has been a growing use of deep learning techniques to detect and classify specific features in images and videos, which, when combined with the use of unmanned aerial vehicles (UAVs) for capturing images, is a potentially useful tool to assist and automate the procedures of visual inspection of facades. This paper aims to perform an analysis of digital image processing for the automatic detection of cracks in ceramic tiles, associated with UAVs or drones, which could potentially bring benefits (time, cost and safety) to the diagnostic process. The research results demonstrated the technical feasibility of detecting cracks through PDI techniques, however, the procedure is a complex process when there is high variation in the study images. At the same time, even in case of a limiting scenario such as the lack of public datasets for the problem, this research project still managed to develop a simple and efficient methodology to deal with the issue proposed.