Method and system for identifying crew overboard in adverse weather
By combining dual-mode camera equipment and an improved U-Net model, the problems of high false alarm and high false alarm rates in crew member overboard identification under severe weather conditions have been solved. This has enabled high-precision, low-false-alarm-rate overboard target identification and rapid emergency response, ensuring crew safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-17
AI Technical Summary
In severe weather conditions, traditional cameras and image recognition algorithms struggle to effectively identify crew members falling into the water, resulting in high rates of missed reports, false alarms, and slow response times, which seriously threaten the safety of crew members.
The system uses dual-mode camera equipment to acquire water surface images in real time, combines preprocessing techniques to remove noise and enhance illumination, uses an improved U-Net model for target recognition, and uses a feature library to make decisions on water-falling events and trigger an alarm mechanism.
It achieved high-precision, low-false-alarm-rate identification of targets falling into the water under severe weather conditions, improving emergency response speed and ensuring crew safety.
Smart Images
Figure CN122416362A_ABST