基于三维模型重投影的结构缺陷识别装置及识别方法
By using 3D model reprojection and small sample recognition technology, the images of UAV inspections are unified onto the surface of the 3D structural model for comparison, which solves the problems of geometric consistency and environmental interference across UAV inspections, and realizes high-precision, robust structural defect identification and digital management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN UNIV
- Filing Date
- 2026-06-17
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, when UAVs inspect large structures, there are problems such as poor geometric consistency across flights, scarce defect samples, severe interference from complex environments, and lack of spatial correlation in the identification results. These problems result in low accuracy and poor robustness in structural defect identification, making it difficult to achieve efficient and digital management.
By acquiring a three-dimensional model of the structure and a health baseline image, and using pose restoration and reprojection techniques, the inspection images are unified onto the surface of the three-dimensional model for comparison. Combined with a small sample recognition model, the precise location and type of defects can be identified.
It achieves high-precision structural defect identification, improves automation and digitalization levels, can stably identify minute defects in complex environments, and directly associates the identification results with three-dimensional spatial location, supporting the digital management of infrastructure.
Smart Images
Figure CN122415618A_ABST