What is the difference between the cameras that support ONVIF configuration for analytical modules and those that do not?

How do I know if my camera can be configured over ONVIF?

Cameras that support ONVIF configuration 

Cameras that DO NOT support ONVIF configuration 

How do I know if my camera can be configured over ONVIF?

After you integrate the camera using either the P&P or ONVIF method the Info Tab of the camera properties window will show you which modules are available for the ONVIF configuration. Different camera models will have different levels of ONVIF compliance so in some cases all modules of the camera can be Available for ONVIF configuration, and for other camera models, only some or even none of the modules will be available for ONVIF configuration.

 

Cameras that support ONVIF configuration 

For cameras that do support the configuration of the analytical modules over ONVIF, we recommend using our built-in tools to configure detection zones and sensitivity. Besides being a more convenient way, configuring camera-side analytics through the Analytics tab will ensure that the configurations on the camera and in our system are the same. One of the reasons why it is important to have the same configuration on the camera and on the Analytics tab is to ensure that when the AI is looking for objects the AI will be looking in the correct areas of the image. To learn more about the Object Detection mechanics please visit this page 

When the Admin portal is used to configure the motion detection on the camera the AI would take into consideration the selected "active" area and only when the object is at least 25% inside of the selected motion detection zone the object would be reported.

In the screenshot below, you will be seeing that Line Crossing and Intrusion Detection are greyed out. This is because the modules are supported over the ONVIF but are Disabled at the moment for this particular camera but can be enabled at any time if you choose to do so. If you enable a different analytical module on the camera make sure to adjust the Object detection rule/rules to make sure that the AI is triggered with the correct camera module.

Cameras that DO NOT support ONVIF configuration 

Cameras that are not fully supported over the ONVIF can still be added using both P&P and ONVIF methods but for such cameras, either some or all analytical modules will be not available for ONVIF configuration. For such cameras, the analytic modules on the camera have to be configured from the Camera's native Web UI.

For the cameras with edge analytics that is not available for the ONVIF configuration the whole frame would be considered as an active detection zone at all times and any object anywhere on the image that is moving would be reported. The possibility of limiting the areas of the image to a specific region to reduce the false positive is reduced in cameras that do not have full ONVIF compliance of edge analytics. 

Additionally, only Motion detection can be used as a trigger for such cameras thus when configuring the camera to detect events make sure you are using Motion detection only, otherwise, the AI won't be triggered and thus Object detection won't work.

To configure the motion detector on such cameras you will have to access the camera web UI directly. For the cameras that were added using P&P option you can use Generate URL button from the Troubleshooting tab. This will create a temporary link that will allow you a direct access to the cameras web UI for up to 60 minutes time, if you need longer access simply generate a new link once the original expires.

For cameras that were added using the ONVIF option, you can use the HTTP access URL from the Network tab of the camera properties to access the camera's web UI directly. 

Best practice for ONVIF-compliant cameras

• Configure the camera-side analytics modules using the Analytics tab in the Admin portal
• Configure desired zones for detection to limit false objects in the areas that are outside of the region of interest
• Set sensitivity to a high enough level to ensure that camera is not missing any important events but not too high so that the camera is not sending a continuous event. The optimal sensitivity value will depend on the camera model and in some cases could be as low as 10% while for other models even 85% could be too low and 90% or above have to be used. The only way to determine this is to experiment. 
• Configure the Object detection rule to use the detector that you have configured on the camera

Best practice for non-compliant cameras

• Configure the motion detection from the camera's native Web UI as per requirement.
• Set sensitivity to a high enough level to ensure that camera is not missing any important events but not too high so that the camera is not sending a continuous event. The optimal sensitivity value will depend on the camera model and in some cases could be as low as 10% while for other models even 85% could be too low and 90% or above have to be used. The only way to determine this is to experiment. 
• Configure the Object detection rule to use Motion detection as a trigger