Method and system for identification and classification of low, slow and small flight targets based on multi-modal large model
By fusing data from multiple sensors using a multimodal large model, the problem of insufficient accuracy in identifying low, slow, and small flying targets in existing technologies has been solved, achieving high-precision detection and real-time early warning functions, and improving the monitoring capability for low, slow, and small flying targets.
CN119513719BActive Publication Date: 2025-11-25SHENYANG INST OF COMPUTING TECH CO LTD THE CHINESE ACAD OF SCI
Patent Information
- Application Number
- CN202411526978.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-30
- Publication Date
- 2025-11-25
- Estimated Expiration
- 2044-10-30
Smart Images
Figure CN119513719B_ABST
Abstract
The present application relates to a kind of low, slow and small flight target identification and classification method and system based on multi-modal large model, aims at the monitoring and identification problem of low, slow and small flight target in complex scene.This application fuses a variety of sensor data, including visible light image, infrared image, radar signal, sound feature and radio signal, processes and fuses the data of different modalities using multi-modal large model, extracts joint features, realizes the efficient detection and classification of low, slow and small target, and has automatic warning function.This application has the advantages of real-time response, automatic processing and high recognition accuracy, and is suitable for airport, military base, city security and other low, slow and small flight target defense scenes.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Patent Citations
Comprehensive detection and identification method and system for aerial flying object
CN115620176A
Foreign matter expelling method and system
CN118827721A