The Object Detector can also detect whether an object has been moved or removed from its original position. The detection algorithm of Object Detector is highly sensitive and can even detect small objects (5×5 px) such as lighters or pocket knives.
Object Detector is designed to detect objects with a predeﬁned minimum object size. If detected objects are not removed after a certain period of time, the application will trigger an alarm. Alarms can also be suppressed when there is motion in the detection area or on the detected object.
Fields of Application
The analysis of Object Detector is based on constant lighting conditions in the applied scene (e.g. in elevators or security gates). When started for the ﬁrst time, the application creates a so called reference image. Every new frame will then be compared to this reference image.
Indoors with changing lighting conditions (e.g. museums, laboratories), Object Detector is learning the diﬀerent possibilities (light on/oﬀ) and will trigger an alarm if lasting changes are detected.
Outdoors, Object Detector is learning the scene independently w.r.t. lighting conditions. In comparison to indoor areas, this mode uses edges to detect objects.