基于动态底库与残差分析的积水检测方法、介质及设备
By constructing a dynamic baseline and residual analysis-based water accumulation detection method, and utilizing multidimensional context labels and abnormal residual features, the method solves the problems of insufficient generalization ability and high false alarm rate in existing water accumulation detection technologies, and achieves high accuracy and stability detection in complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING QIDAISONG TECH CO LTD
- Filing Date
- 2026-05-08
- Publication Date
- 2026-07-17
AI Technical Summary
Existing water accumulation detection methods lack generalization ability in complex and ever-changing real-world monitoring environments, have a high false alarm rate, and lack effective utilization of spatiotemporal prior information, resulting in fragmented and unstable detection results.
A water accumulation detection method based on a dynamic baseline is constructed. Historical image data is collected through visual perception nodes to build a normal feature baseline. Multi-dimensional context label indexing is used to finely characterize the range of visual feature changes under different environmental scenarios, remove interference from normal environments, and combine single-frame abnormal residual features and abnormal water accumulation evolution patterns for detection.
It significantly reduces the false alarm rate, improves the accuracy and robustness of detection, and can accurately identify water accumulation in complex environments, adapting to changes in different scenarios.
Smart Images
Figure CN122157170B_ABST