Channel occupancy detection camera
Product Category：
It solves the problem that the "item left guard" in the traditional sense does not have a high false positive rate due to deep learning, and the disadvantage that simple deep learning is difficult to adapt to a variety of scenarios, resulting in a low recognition rate. After combining the original calibration of a large number of channel scenarios, the actual on-site background after installation is used for "secondary calibration", which is equivalent to the role of secondary calibration. On this basis, it is identified through deep learning and powerful AI computing chips.
Order Hotline：139-1551-1983
Product Details
Consultation Hotline：139-1551-1983
Application scenarios
It supports channel occupancy detection through AI channel detection and recognition algorithm, and is reviewed by the platform or NVR through images or videos to realize the visual detection of evacuation channels, safety exits, fire truck channels, etc.
Intelligent parameters
Lens: F1.6, 4mm, 5MP
Field of view: horizontal 90°, vertical 48°
Supplementary light: 4 white lights + 8 infrared lights
The farthest light supply distance: 30m
Minimum recognition pixel: 130*130
Stable duration of the scene: ≥10s
Detection logic: When there is a change in the picture (there is an obstacle, or the existing item is moved), the changed scene is stable (no people or other objects move) more than 10s, that is, the judgment channel is occupied.
Detection mode: both daytime, infrared and warm light mode
Linkage alarm: Support linkage alarm
Detection area: specified area, default full-screen detection
Control time: support custom control time
Trigger method: picture, video trigger
Online Message
