| I&T Solution |
AI Video Analytic System for Marina Management
(REF: S-1873) |
| Trial Project |
|
| Solution Feature |
- The integration of deep learning AI for video analytics facilitates real-time classification of objects within the marine environment. This capability allows for the immediate identification of vessels and other objects enhancing situational awareness and operational safety.
- By employing advanced algorithms, the system processes video footage to detect and interpret the wake generated by the vessel, allowing for accurate speed calculations in real-time. This system can detect violations such as vessels exceeding speed limits.
- A flexible alert system can notify personnel if a vessel remains parked outside designated mooring areas for a designated period. Users can adjust notification parameters based on preferences, ensuring tailored oversight that meets specific operational needs and enhances security protocols.
- Utilizing Edge AI computing allows for real-time data processing on-site. This reduces latency and mitigate data privacy and security issues by processing and storing sensitive information locally, rather than relying on centralized cloud servers.
- Using SMS and email to send notification. Developing an alert system web-based application allows users to access real-time video footage and analytics. Features include distance measurement between vessels and the coast, and highlighting significant observations from VA, providing comprehensive insights into marine operations.
|
| Trial Application and Expected Outcome |
- Two locations along the selected typhoon shelter will be located with electricity supply and a good top-down view for setting up the video content analyser.
- The Video Content Analyzer comprises an IP camera with zoom capabilities, an AI edge computer, and tailored software. Video analytics are conducted solely on the local edge computer, ensuring that only quantitative results are transmitted to the cloud for web-based presentation, thereby safeguarding privacy concerns.
- The system can track vessel locations and movements in real-time, detecting violations and anomalies like speed limit breaches or obstructions. An alert system notifies security if a vessel remains outside its mooring area for too long, with customizable notification parameters.
- We will meticulously gather authentic data from video samples, subsequently refining our model to attain an impressive accuracy rate of 90-95%. This iterative process will ensure that our analytics are both precise and reliable.
- The web-based application provides real-time analysis and video footage, measuring distances between vessels and the coast, and highlighting key observations. It summarizes vessel counts and shelter occupancy, with email and SMS notifications for alerts, daily overviews, and statistics.
|
| Info on I&T Solution Provider |
| Solution Provider | : | Sebit Company Limited | | Address | : | Unit 316B, 3/F, Enterprise Place, Hong Kong Science Park, Shatin, NT | | Contact Person | : | NG YIU WING |
| Position | : | Director | | Tel | : | 39052033 | | Email | : |
wingng@sebit.world | | Webpage | : | https://sebit.world |
|
