Identification of Buried Pipes Using Thermal Images and Data Mining

Abstract

Knowledge of buried infrastructure is essential for management and decision-making in water utility companies. To pinpoint the benefits of thermography, a case study was conducted, namely, thermal images were taken under various conditions in a particular area containing a buried pipe. Thermal images were subsequently analysed, an image database supplemented with environmental data generated, and a data mining method applied. The results generated this way complement the raw images and reveal additional relationships regarding the visibility of the infrastructure, and the possibility of making promising improvements in infrastructure identification, which can be applied in subsequent work in other applications.