一种基于深度学习的边坡裂缝检测方法及系统
By combining a slope structure constraint model with deep learning, the problems of inaccurate crack candidate region localization and insufficient continuity in slope crack detection are solved, achieving higher recognition accuracy and stability, and making it suitable for slope detection in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHANGCHUN GOLD DESIGN INST
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing slope crack detection methods suffer from inaccurate crack candidate area localization and insufficient crack continuity, leading to misidentification or omission, making it difficult to generate continuous crack identification information suitable for engineering interpretation.
Using multi-source inspection image data, a slope structure constraint model is constructed through a geometric texture feature layer and a thermal anomaly feature layer. Combining a strip target pattern recognition algorithm and deep learning, structural constraint analysis and continuity guidance are performed on the crack candidate region to generate an initial crack identification region. Semantic consistency verification and category determination are then performed.
It improves the accuracy and stability of slope crack identification, enhances applicability to complex backgrounds, reduces interference, and improves the continuous extraction capability of slender cracks.
Smart Images
Figure CN122115449B_ABST