UAV-Assisted quantitative infrared thermography procedure for thermal measurements in building inspections

The increasing adoption of unmanned aerial vehicles (UAVs) equipped with infrared thermography (IRT) sensors offers new opportunities for rapid and non-invasive assessment of building envelopes. However, the transition from qualitative thermal anomaly detection to reliable quantitative measurements remains limited by the lack of standardized procedures for controlling acquisition conditions, radiometric corrections, and validation strategies. This study addresses this gap by developing and experimentally validating a UAV-assisted quantitative IRT procedure for acquiring external apparent surface temperature data on existing buildings under controlled field conditions. The proposed method combines preliminary terrestrial IRT surveys, environmental monitoring, contact surface-temperature measurements, repeated UAV-based thermal acquisitions for estimating reflected apparent temperature, and controlled post-processing based on documented spectral emissivity, target distance, relative humidity, and region of interest (ROI) definition. The methodology was experimentally evaluated through six case studies in Milan (Italy), using terrestrial IRT measurements as reference data and comparative measurements acquired with two different thermal imaging cameras mounted on UAVs, together with contact temperature sensors. The results show that UAV-derived apparent surface temperature values can be considered reliable only when a tightly controlled measurement workflow is adopted. Target distance, gain mode, reflected apparent temperature estimation, ROI geometry, and software averaging method significantly affect the resulting temperature values. The High Gain mode produced acceptable or borderline deviations in several short-distance acquisitions, whereas the Low Gain mode generated unacceptable deviations for the tested range of surface temperatures of the building envelope. The study transferability the acquisition, correction, and validation steps required to improve the reproducibility and transferability of UAV-based IRT in apparent surface temperature measurements for quantitative building-diagnostics applications.