To mitigate the safety hazards posed by the frequent detachment of facade tiles, this study summarized the causes of these defects, identifying hollowing defects as precursors to exterior wall tile detachment. Using an Unmanned Aerial Vehicle (UAV) equipped with infrared thermal imaging camera, the study conducted laboratory tests for hollowing detection. The optimal observation attitude of the UAV was investigated. The impact of defect characteristic parameters on identification accuracy and the effect of UAV rotor operation on the temperature of external walls were evaluated. Additionally, a temperature difference threshold was proposed for the identification of hollowing in exterior wall facade tiles. The study indicated that optimal observation occurs when the UAV is 2 to 3 meters from the external wall, with a vertical angle of -30° to 30° and a horizontal angle of -15° to 15°. The drone rotors increase the cooling rate of external walls by 10%. Recognition improves with hollowing of larger sizes, shallower depths, and greater thicknesses. Hollowing defects in black and red tiles exhibited higher temperatures, while those in yellowish tiles exhibited lower temperatures. Consequently, an outdoor test was conducted to verify the efficacy of the proposed detection method by comparing it with visible light image recognition, which provides a novel threshold for the expeditious identification of hollowing defects in facade tiles.