Kipod has demonstrated a new cloud platform of the State Monitoring System that can be considered a useful tool for Smart City project. A global video surveillance system with an artificial intelligence engine shall be deployed across the Republic of Belarus and connect about 400,000 cameras to 100,000 users from various security services.
Kipod has demonstrated a new cloud platform of the State Monitoring System that can be considered a useful tool for Smart City project. A global video surveillance system with an artificial intelligence engine shall be deployed across the Republic of Belarus and connect about 400,000 cameras to 100,000 users from various security services.
The State Monitoring System for Public Safety (SMS) provides such carrier class features as:
- instant video content search across big data from any web browser and mobile device
- open source software infrastructure (Linux, Kubernetes, Cassandra and Ceph)
- scalability beyond 1M users, 1M cameras and 1M GB (1 exabyte) video storage
- tier 3 availability with full redundancy and background software updates
Kipod demonstrated live video monitoring and forensic tools of the cloud platform in the live urban environment. The system includes technologies, based on deep-learning neural networks, such as face recognition, number plate recognition and vehicle identification, human behavior video analytics and sound recognition (audio analytics).
Kipod cloud video analytics capabilities increase the operators’ work efficiency by prioritizing their time and effort, improve visitor screening and enhance security by reacting timely to alarm monitor. For example, the platform will alert its users whenever an unexpected behavior or a certain face appears in a camera view.
One of the Kipod modules is audio analytics module which can detect a scream, a gunshot and a sound of glass breaking while ignoring a background noise. This feature improves overall system efficiency by being able to work in any conditions and not being affected by weather and time of the day.
Kipod number-plate recognition (NPR) feature allows to conduct quick and efficient search within the archive under few minutes and easily identify vehicle of interest and unwanted movement. It also enables the user to create number plate lists, and in case the vehicle is detected in this list, the user will be notified via alarm monitor.
Dr Nikolai Ptitsyn, Kipod' CEO, said: “The Kipod platform revolutionises urban CCTV using the cutting edge cloud and machine learning architecture. Authorized users from different organizations can quickly search millions of gigabytes of video content in a very simple way, like they search the Internet.”
This system will allow to take government security on the whole new level and enable the operators to timely react on alarm events and within few minutes search data archive to retrieve necessary information.