icetana Limited
aisetana
How anomaly detection works
Icetana's anomaly detection not only detects people, but also vehicles, motorcycles, fire, smoke, and abandoned luggage.
Icetana's autonomous learning AI can detect a wide range of "unusual" conditions for each camera. It can detect events that are difficult to predict or define with an accuracy similar to that of patrols or sentries, enabling the necessary response to be taken in real time.
In addition, icetana can accurately detect specific conditions in specific areas, such as falls, obstructions, and intrusions.
By using both autonomous learning and rule-based technologies depending on the requirements, we can achieve more efficient and sophisticated facility security.
1. Unusual “position”
An abnormality is detected when a person, vehicle, or motorcycle is in an area where it should not be.
<Detection example>
・Staying due to eating/drinking/photography or nuisance
・Dangerous acts (children's mischief, delivery on different routes)
・Intrusion into partitions/plants
・Smoking, cigarette littering
・Movement overcoming measures
・ Violation injection, bicycle parking
・Passengers on roadways and vehicles on sidewalks
2. Unusual “number”
Abnormality is detected when the number of people, vehicles, and motorcycles is abnormally larger or smaller than usual.
<Detection example>
・A nuisance such as hanging out, drinking alcohol, or taking photos without permission
·Traffic jam
・Violation of number limit
・Protest
・Using an emergency exit
3. Unusual “speed”
Detects people, vehicles, and motorcycles that move faster than normal as anomalies.
<Detection example>
・Dangerous work (such as unreasonably fast transportation)
·skateboard
・Vehicle slip
・Cars that are speeding
・Bicycle/motorcycle traffic violations
4. Unusual conditions “by time of day”
Anomalies are detected when people, vehicles, or motorcycles are present during an unexpected time period.
<Detection example>
・Nighttime hanging out/drinking
・Intrusion after closing
・Use after hours
・Intrusion into the partition area
5. Unusual “movement in direction”
When a person, vehicle, or two-wheeled vehicle moves in an unusual direction, it is detected as an anomaly.
<Detection example>
・Reversing on the escalator
・Reversing the escalator
・Vehicle/bicycle reverse/slip
6. Human fall
An abnormality is detected when a person lies on the floor for 3 seconds or more.
<Detection example>
・Falling/injured person
・Maintenance worker lying face down
・The child is lying down
Violent behavior leading to falls
7. Generation of fire and smoke
An abnormality is detected when fire or smoke is detected.
<Detection example>
・Ignition and smoke from equipment
8. Retention
Detects when people are staying.
<Detection example>
・A person who stays in the same place for a certain period of time
・Crowds at the entrance/exit of the facility
・Gathering of smokers
9. Person/vehicle count
Displays a count-up of the number of people and vehicles detected per hour/day for each camera.
<Example of use>
・Understanding congestion times and locations
・Understanding traffic volume around the facility
・Obtain marketing data and provide it to tenants
10.Heatmap
Displays the relative value distribution of the residence time/number of people/vehicles/bicycles within the field of view of each camera.
<Example of use>
・Understanding traffic conditions and key areas
・Understanding residence time
・Review and strengthen security system
・Improvement of facility layout
11. Leaving luggage unattended
Detects when luggage (suitcase or backpack) is left unattended for a certain period of time.
<Example of use>
・Prevention of theft
・Reducing potential risks by identifying suspicious objects
Five functions of icetana beyond just anomaly detection
By utilizing icetana, in addition to anomaly detection as a security solution, you can use five functions - "data analysis," "forensics," "vehicle number detection," "face recognition," and "icetana GPT" - in a single system.
Inquiry
If you have any questions about icetana's anomaly detection or other features, please feel free to contact us.