Application of Deep Learning and Unmanned Aerial Vehicle on Building Maintenance

Several natural and human factors are responsible for the defacement of external walls tiles buildings, related deterioration can be a public safety hazard. Therefore, active building maintenance repair processes essential ensuring sustainability. However, conventional inspection methods time-, cost-, labor-intensive processes. herein, this study proposes convolutional neural network (CNN) model image-based automated detection localization key defects (efflorescence, spalling, cracking, defacement). Based on pretrained CNN VGG-16 classifier, applies class activation mapping object localization. After identifying its limitations in real-life applications, determined model’s robustness ability to accurately detect localize wall buildings. For real-time localization, applied by using mobile devices drones. The results show that application deep learning with UAV effectively various kinds improve efficiency.