Modelamiento térmico 3D aplicado a la inspección de infraestructuras usando vehículos aéreos no tripulados

Abstract

There is a growing trend in the use of unmanned aerial vehicles (UAVs) to perform infrastructure inspection work as it provides unique advantages compared to traditional inspection methods, because it has proven to be a non-invasive, flexible, accurate tool, efficient and low cost compared to the use of other conventional tools and methods, which allow an evident saving in the costs of information acquisition, as well as the reduction of probability of human accidents in the inspections, however with respect to the use of sensors thermal aboard UAV there are large gaps that still need to be investigated in greater detail, such as biases in the automation of data processing, development of methodologies depending on the type of infrastructure, others. In this study, the photogrammetric reconstruction of a 3D model based on thermal aerial images obtained with a thermal camera on board a UAV is carried out. From this 3D thermal model, the visual inspection and detection of thermal anomalies was carried out. El Ingenio heritage building, detecting 107 anomalies of which 62 anomalies were true positive, 20 as false positives and 25 anomalies were identified as false negatives, finally the research concluded an accuracy index of 75.61% and 71.26% for the repetition rate.