Building health monitoring with different thermal infrared cameras

Building health monitoring is an essential issue in maintaining building sustainability. Thermography has been widely employed for monitoring buildings. Thermography records the surface temperatures of a target. Those areas with abnormal surface temperatures with their neighborhoods can be treated as defects. Image segmentation groups those pixels with similar surface temperatures such that the recorded thermography can comprise several segmented regions. Those segmented regions offer an essential clue for defect detection. Recently, a thermal camera has been installed on an unmanned aerial vehicle (UAV) to collect the surface temperatures of a building. Those collected thermal infrared images are analyzed with image segmentation. With the segmented regions, the potential defects can be identified. On the other hand, a thermal camera installed on the ground is used to record a series of thermal infrared images. The series of thermal infrared images were analyzed using robust principal component analysis (RPCA) to project the given data onto a low-dimension and feature space. The first image extracted from the low-dimension space inherits the significant properties of the recorded images such that the extracted image can be segmented to illustrate the defects. Two processed results are compared. UAVs provide an efficient way to monitor the building's health conditions, and periodic ground observations offer a stable way to monitor the building. Both ways provide an efficient and robust method to monitor the health conditions of a building.