Related Experiment Video
Updated: Jan 9, 2026
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
43.5K
UAV and Deep Learning for Automated Detection and Visualization of Façade Defects in Existing Residential Buildings
Yue Fan1, Jinghua Mai1, Fei Xue1
1State Key Laboratory of Subtropical Building and Urban Science, Center for Human-Oriented Environment and Sustainable Design, School of Architecture and Urban Planning, Shenzhen University, Shenzhen 518060, China.
Sensors (Basel, Switzerland)
|December 11, 2025
Summary
This study introduces an automated framework using drones and AI for detecting building façade defects like cracks and moisture. It offers an efficient solution for urban building safety management.
Area of Science:
- Urban planning and infrastructure management
- Artificial intelligence in civil engineering
- Remote sensing and geospatial analysis
Background:
- Accelerated urbanization leads to prominent façade defects in residential buildings, threatening structural safety and living quality.
- Traditional manual inspection methods in high-density urban areas like Shenzhen are inefficient and prone to errors.
- Existing methods lack comprehensive defect detection and efficient data visualization for large-scale building management.
Purpose of the Study:
- To develop and validate an integrated framework for automated façade defect detection using Unmanned Aerial Vehicle (UAV) technology.
- To combine visible-light and thermal infrared imaging with deep learning for synergistic defect identification.
- To enable real-time 3D visualization and parametric analysis of detected façade defects for improved building safety management.
Main Methods:
- Utilized Unmanned Aerial Vehicle (UAV)-based visible-light and thermal infrared imaging for data acquisition.
- Employed deep learning algorithms (Knet-based model) for automated crack and moisture intrusion detection.
- Developed a Grasshopper-integrated mapping tool for parametric 3D visualization of defect information.
Main Results:
- Fusion of visible and thermal infrared images effectively identified cracks and moisture intrusion defects.
- Optimized shooting distances were determined: 5-10 m for low-rise and 20-25 m for high-rise buildings.
- The Knet model achieved 87.86% mIoU for crack detection and 79.05% for leakage detection.
Conclusions:
- The proposed UAV-based framework provides an efficient and accurate technical solution for city-wide building safety management, especially in densely populated areas.
- This integrated approach enhances the detection of critical façade defects and facilitates data-driven maintenance strategies.
- The framework supports the development of automated building maintenance systems and smart city infrastructures.
Related Concept Videos
01:20Masonry Curtain Walls
1.6K
Masonry curtain walls employ brick or stone veneers supported by the building's structure to form an external cladding system that is both aesthetically appealing and functional. These walls are erected through two principal techniques, first by traditional layering of masonry units and second by using prefabricated panels. Traditional construction relies on steel shelf angles attached to the spandrel beam for support, with high-bond mortars ensuring secure attachment of masonry veneer...
1.6K
01:27Movement Joints in Buildings
311
Movement joints in buildings are essential design elements that accommodate inevitable motions caused by various factors such as temperature changes, moisture content variations, and structural deflections. These motions, if not considered in design and construction, can lead to unsightly or dangerous damage. Movement joints are incorporated in different forms to manage these stresses and allow materials to move without causing distress.
The simplest type of movement joints, working joints, are...
The simplest type of movement joints, working joints, are...
311