AI Features

Loitering Detection

Loitering Detection

Detect people, cars, and bicycles loitering in specified areas. This can enhance security by detecting suspicious behavior.

It detects when an object enters a monitored area and remains there for a specified period, subsequently triggering an alarm. This application is designed for security monitoring, eliminating the need for 24/7 recording or constant real-time monitoring by an administrator. Instead, it sends immediate alerts in response to specific events, ensuring effective protection of critical assets in buildings, perimeters, and restricted access areas, and enabling rapid incident response. Detection targets can be configured for specific objects, such as people, cars, and bicycles. The application is designed to minimize false alarms by considering factors such as reflections from stationary objects, vibrations, and the influence of nearby objects. Additionally, even ambiguously shaped objects are detected, and a filtering process is applied to trigger only genuine events, minimizing false alarms even in complex environments.

Camera Installation Requirements and Minimum Object Size

Camera Installation Environment
  • Indoor(O)
  • Outdoor(O)
Supported Camera Type
  • IP Camera(O)
  • Fisheye camera (limited functionality)
  • Thermal Camera (X)
Input Resolution
  • Min. 640 x 360 ~ max. 4K
Input Frame Rate
  • Min. 3 fps

Minimum size: 1280 x 720 resolution or higher
  • ① Day
    [Person] W 1.5%, H 4%
    [Car] W 1.5%, H 2.5%
    [Bicycle] W 2%, H 4%
  • ② Night
    [Person] W 2%, H 6%
    [Car] W 2.5%, H 4.5%
    [Bicycle] W 3%, W 5%
Minimum size: from 640 x 360 to 1280 x 720 or smaller
  • ① Day
    [Person] W 2.5%, H 6.5%
    [Car] W 2.5%, H 3.5%
    [Bicycle] W 2%, H 4%
  • ② Night
    [Person] W 2.5%, H 6.5%
    [Car] W 3.5%, H 5.5%
    [Bicycle] W 3%, H 5%
Fisheye Video
  • [Person] W 4%, H 5%

Videos and tutorials

Loitering Detection