Issues facing the security scene
In large-scale facilities, there are many security cameras and the duties of the security guards who monitor them are diverse, making it difficult to detect abnormalities such as accidents or injuries or illness in real time through surveillance security. Even if the video can be used, in most cases, it is only necessary to respond after the fact, such as checking the recording. Therefore, in order to strengthen security, it is necessary to increase the resources for monitors, patrols, and sentry security, and although the introduction of mechanical security and AI has been considered, the implementation of machine security and AI is limited, and the role of security (various It is difficult to be responsible for detecting abnormalities and signs, and it takes time and effort to operate.
As a result, many companies may have decided that they have no choice but to continue with manual security.
icetana solves
We detect all sorts of anomalies and their signs from security camera footage, realizing labor savings in surveillance operations, patrols, and security guards. Furthermore, since it can support a wide range of requirements and is easy to operate, it can be reliably put into practical use.
Solution brief
icetana is a security camera support system that utilizes AI image analysis technology, and supports the discovery of any abnormalities from camera footage.
``Unusual conditions'' are detected and displayed to the monitoring staff, who can check the detection contents of icetana, judge the situation, and then decide whether or not to take action.
Since it can be used in conjunction with existing security camera systems, security camera monitoring enables efficient and wide-ranging situational awareness.
Mechanism and operation of anomaly detection
Normal state learning
・Learn the "position, concentration, direction and speed of movement" of a collection of pixels "for each time period"
・Specific things and events such as people, objects, accidents, and incidents are not identified.
・Learn by distinguishing between weekdays and holidays
Detection during operation
- Quantifies the difference from normal conditions, detects images that exceed a predetermined threshold, and displays only those images to the supervisor.
・Always automatically update and detect the last two weeks as normal state data
・Since it detects anything that is even slightly different from normal conditions, most images are displayed without any problems that do not require any action.
Monitoring and operation method
・Monitors check only the images detected by icetana and decide whether or not to take action.
・If no action is required, leave it as is.
・If response is required, respond in accordance with the rules of the site.
Features
1. You can discover all kinds of abnormalities and their signs.
Able to respond to unusual movements such as unexpected events.
2. No need to worry about system operation
The status of the last two weeks is automatically updated as the normal status, and abnormalities are detected with higher accuracy. There is no need to set detailed rules to detect specific events.
3. Highly scalable
There is no need for extensive preparation or expense to change the applicable cameras or increase the number of cameras, and it can be used on a scale of hundreds of cameras.
Contact Us
We introduced the icetana solution. What did you think?
If you are interested, please contact us using the button below.
Latest Information
Case Study
Blog
Document List