AI-INTRUSION-PRO



The performance is not adequate, what else can I configure?



WARNING: Please read the disclaimer to make sure that all of the installation constraints have been fulfilled!

Go back


Installation Constraints

Camera positioning should take into proper consideration the distance from the scene and the object whose detection is desired. Indeed, to enable a proper detection, the object of interest should at least measure 10x10 px. Intrusion

Choosing the right lense also depends on the resolution at which it possible to elaborate the video stream; the followings are instructions on this matter. Go back


Sterile area sensor configuration

Aree Sterili In the drop-down list on the left, by selecting "Algorithm Parameters" and then “Intrusion Sensor” you can add, remove and edit sterile areas within the scene framed by the camera. Specifically, by clicking on “Add sensor” you can draw the area of interest directly on the live image on the left. The area of interest is shown as a polygon with any number of sides. In the field “Sensor Name” specify the name associated to the sensor that will be used also when notifying events to various receivers.

Moreover, it is also possible to set the following parameters:

  • Sensor Type: you can choose between two types of sensors:

    • Impulsive: generates a single event for the whole duration of the intrusion;
    • By levels: generates events of beginning of intrusion, persistence of intrusion (every “Inhibition” seconds) and end of intrusion.
  • Time of end of intrusion: after this timespan of seconds, provided that the sensor by levels does not detect intrusions anymore, the system sends an event of end of intrusion. Go back

Line crossing sensor configuration

Attraversamento Linee In the drop-down list on the left, by selecting "Algorithm Parameters" and then “Line Crossing” you can add, remove and edit sensors of line crossing within the scene framed by the camera. Specifically, by clicking on “Add sensor” you can draw a virtual line directly on the image acquired by the camera. The virtual line is represented by a segment; there is an arrow starting in the centre of this segment that shows the direction of crossing detected by the sensor (crossings in the opposite directions are not indicated).
In the field “Sensor Name” specify the name associated to the sensor that will be used also when notifying events to various receivers.

Moreover, it is also possible to set the following parameters::

  • Crossing line pre confidence: confidence associated to the activation of pre-alarm on the object
  • Crossing line post confidence: confidence associated to the activation of the alarm on the object in pre-alarm state
  • Crossing line pre latency: latency time in seconds of an object in the scene before crossing the line (pre-alarm)
  • Crossing line post latency: latency time in seconds of an object in a pre-alarm state

Go back

Multiple line crossing sensor configuration

Attraversamento Linee Multiple Configuration of application AI-INTRUSION PRO version acquires all the different functions already existing in AI-INTRUSION:

  • Intrusion sensor
  • Line crossing

In addition to this, there is also “Aggregated lines crossing”. Thus, the following are the specific parameters for this section. Here you can define a path of lines crossed within the scene. After the crossing of all of the lines a PRO version INTRUSION event is generated. To define the path add an aggregated lines crossing sensor by clicking on the devoted section.

After this step, you should also:

  1. Add the lines of the desired path by using “Line crossing sensor
  2. Specify the name of the aggregated sensor that will be shown in the event and the minimum time for crossing each line. The second parameter, in seconds, specifies the maximum time for crossing two consecutive lines in the sequence created by the user.

Other parameters to configure are the following:

  • Crossing line pre confidence: confidence associated to the activation of pre-alarm on the object
  • Crossing line post confidence: confidence associated to the activation of the alarm on the object in pre-alarm state
  • Crossing line pre latency: latency time in seconds of an object in the scene before crossing the line (pre-alarm)
  • Crossing line post latency: latency time in seconds of an object in a pre-alarm state A numeric ID is associated to each line of the path; this ID shows the desired sequence of activation (as such, the specific order of the line is a crucial information).