大坝混凝土缺陷空地巡检系统及其方法
By constructing a defect sample database of high dams and large reservoirs and performing multi-source analysis and fusion, a lightweight defect target detection model was trained, solving the problem of manual reliance in dam inspection, achieving efficient defect identification and quantification, and improving the accuracy and efficiency of inspection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HOHAI UNIV
- Filing Date
- 2024-11-16
- Publication Date
- 2026-07-17
AI Technical Summary
The current inspection of concrete defects in dams mainly relies on manual labor. Experience and skill level have a significant impact on the inspection results. Areas that are difficult to reach or where the line of sight is obstructed are often overlooked, leading to the omission of potential defects. The existing equipment and technology are not widely and deeply applied, making it difficult to meet the requirements of modern water conservancy projects for inspection efficiency and accuracy.
A defect sample database for high dams and large reservoirs is constructed. Defect sample enhancement learning is performed through digital image processing models to generate comprehensive defect enhancement samples. Multi-source analysis and fusion are conducted to train an engineering defect target detection model. The model is then lightweighted and deployed to edge devices for inspection.
This improved the efficiency of dam defect inspection, reduced the occurrence of safety hazards, and enabled the rapid identification and quantification of defects.
Smart Images
Figure CN119579539B_ABST