The MxActivitySensor reacts to continuous movements within specific image areas. You can define one or more independent image areas for each image sensor (camera image). Since the MxActivitySensor only reacts to relevant movements, events are triggered with a certain delay after the first appearance of an object. For recording, you can compensate this behavior by adjusting the Recording Time Before Event accordingly. Likewise, the measured activity fades away with a certain delay, which may prompt additional alarms. You can control this behavior by adjusting the Event Dead Time accordingly.
In order to obtain best results, the scene should be well-illuminated, so that the objects that are to be detected show sufficient contrast against the background of the image. In addition, you should choose the camera image in such a way, that the camera can follow objects within the detection area for several seconds.
Note that you can create different image analysis events, which react to different detection areas. In addition, it is also possible to detect only specific directions of movement.
Parameter |
Description |
---|---|
Detect Fast Motion |
By activating this checkbox, you can improve the detection of fast objects by MxActivitySensor. However, using this option can reduce the recognition results of small objects. |
Activate Noise Improvement |
Improves the detection precision in low-light situations with a high degree of image noise by filtering the noisy pixels from the image information. |
Show Direction Indicator |
Activates the direction indicator that is displayed in the live image. This display combines all active profiles of the MxActivitySensor type. The direction indicator is not available for every display mode. |
Show Live Position Indicator |
This option allows showing the position of the movement that triggered. The position indicator is not available for every display mode. |
Threshold for Thermal Sensors |
Allows setting the threshold of the MxActivitySensor for thermal sensors. The lower the value, the more sensitive the MxActivitySensor will react. This parameter is only available if at least one thermal sensor has been attached. |
Show Detection Area |
To display the detection areas of an MxActivitySensor profile, select the corresponding profile in the Image Analysis Events dialog. |
This extension to the MxActivitySensor provides a method for object detection based on object size. It can thus eliminate false alarms triggered by objects that are too small. For example, you can use this feature to avoid false alarms triggered by animals (cats, dogs, birds, etc.).
The parameters and options in this section are only available if the MxActivitySensor 2.x option is activated.
If several image sensors are attached, you can set different parameters for each sensor.
Using the MxActivitySensor 2.x may slightly reduce the maximum frame rate.
The perspective compensation provides a method for determining the size of an object independent of its distance to the camera. To be able to do so, the camera needs to know the minimum size of "real" objects near the camera and some distance away. The camera software uses the relative position of the windows in the image to determine the distance to the object.
Without perspective compensation, you can set a minimum object size from which the corresponding object triggers.
You will get the best results if the camera is installed in a slightly elevated position so that it is tilted downward by about 10 to 15 degrees.
When defining the object size, make sure that the size windows are tight around the objects. If you want to trigger on humans, for example, draw the windows snugly along the body (from the throat down to the feet, ignore the arms and the head, as wide as the torso).
The perspective compensation is not available for hemispheric sensor modules.
Parameter |
Description |
---|---|
Deactivated Perspective Compensation |
If the perspective compensation is deactivated, only the object size (as indicated by the changing pixels in the image) will be used to trigger an event. The following parameters are available:
|
Activated Perspective Compensation |
If the perspective compensation is activated, the object size to trigger an event will be set twice − once for the object near the camera, once in some distance. The following parameters are available:
|
Rectangle syntax:
Horizontal position, vertical position, width, height
Parameter |
Description |
---|---|
Horizontal position |
Number of pixels measured from the left edge of the visible image to the left border of the area (based on a MEGA image 1280x960). |
Vertical position |
Number of pixels measured from the bottom edge of the visible image to the bottom border of the area (based on a MEGA image 1280x960). |
Width |
Width of the area in pixels (based on a MEGA image 1280x960). |
Height |
Height of the area in pixels (based on a MEGA image 1280x960). |
If you would like to define a new measurement area, open the Image Analysis Events dialog and follow the instructions of the Graphically Entering Image Areas help page.
Every line in the definition box that does not begin with a hash sign (#
, comments) defines one window. The windows can be defined either as rectangles or as polygons.
Rectangle syntax:
[!]image sensor, horizontal position, vertical position, width, height
(optional parameters have square brackets).
0,270,190,100,100
Polygon syntax:
[!]image sensor, poly=<hor. coordinate>x<vert. coordinate>{/<hor. coordinate>x<vert. coordinate>}
(optional parameters have square brackets, parameters that can be used as often as needed have curly brackets).
0,poly=270x190/270x289/369x289/369x190
All window dimensions refer to the entire image sensor.
If the camera image has been mirrored or rotated, the coordinate system of the windows changes accordingly.
Parameter |
Description |
---|---|
! |
Defines an area that is excluded from analysis (this area should always be inside of the area that is to be analyzed). |
Image sensor |
Selects the image sensor (camera image) to use for this window definition:
|
Horizontal position |
Number of pixels measured from the left edge of the visible image to the left border of the window (based on a MEGA image 1280x960). |
Vertical position |
Number of pixels measured from the bottom edge of the visible image to the bottom border of the window (based on a MEGA image 1280x960). |
Width |
Width of the window in pixels (based on a MEGA image 1280x960). |
Height |
Height of the window in pixels (based on a MEGA image 1280x960). |
Horizontal Coordinate |
Number of pixels measured horizontally from the left edge of the visible image to the polygon corner (based on a MEGA image 1280x960). |
Vertical Coordinate |
Number of pixels measured vertically from the bottom edge of the visible image to the polygon corner (based on a MEGA image 1280x960). |
If you would like to define a new or additional window, open the Image Analysis Events dialog and follow the instructions of the Graphically Entering Image Areas help page.
In order to temporarily deactivate windows, place a hash sign (#
) in front of the definition. This window definition will still be saved, but will not be used as detection area. In addition,
comments generally help keep an overview if the window definitions are complex.
This trigger value sets the activity that needs to be measured in the area of analysis before the MxActivitySensor triggers. The higher the value, the slower MxActivitySensor will trigger; using a lower value require less activity before MxActivitySensor triggers. Value range: 0..99.
In order to determine a good value for the Detection Threshold, you can open the Text & Display Settings dialog and activate the corresponding Level Meter (Bar or Scroll Chart; you may have to click on More at the bottom of the dialog to see these options). Select the corresponding Image Analysis: <event name> entry from the Level Meter Data list and click on the Set button. You can now test the settings by having test persons walk through the image, for example.
Activate the directions of movement that will trigger reactions of the camera (e.g., recording).
If this option is activated, all window definitions will be used on both image sensors. This option is only available on Day/Night models.
Activating this option will detect video motion on both image sensors (regardless of which sensor is actually shown on the Live screen). Note that using this option may reduce the maximum frame rate. A window explicitly defined for the left image sensor will remain active, even if you only see the image of the camera's right image sensor. This means that the camera will react on events detected on the left (invisible) image sensor. This option is only available on Dual models.
Click on the Set button to activate your settings and to save them until the next reboot of the camera.
Click on the Factory button to load the factory defaults for this dialog (this button may not be present in all dialogs).
Click on the Restore button to undo your most recent changes that have not been stored in the camera permanently.
Click on the Close button to close the dialog. While closing the dialog, the system checks the entire configuration for changes. If changes are detected, you will be asked if you would like to store the entire configuration permanently.