一种基于多模态场景识别的摄像头控制方法及装置
By using dynamic analysis and feedback loop optimization of multimodal data streams, the problems of poor data coordination and delayed response in multimodal monitoring systems under dynamic abnormal scenarios are solved, achieving more efficient camera control and anomaly identification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 深圳市互通创新科技有限公司
- Filing Date
- 2025-10-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing multimodal monitoring systems suffer from poor data coordination and delayed camera control response in dynamic and abnormal scenarios.
By acquiring multimodal data streams, integrating initial data packets, calculating the modal distribution of the current scene, monitoring data stream speed, identifying abnormal scenes, adjusting acquisition parameters, optimizing the transmission queue, generating control commands, establishing feedback loops, optimizing the multimodal fusion sequence, calculating the abnormal risk level, and generating final control commands.
It enables dynamic correlation analysis of multimodal data, improves the transmission efficiency of key data and the timeliness of camera control in abnormal scenarios, reduces the risk of missed and false judgments in monitoring, and provides more reliable intelligent monitoring and industrial automation support.
Smart Images
Figure CN121309961B_ABST