Motion detection on only part of the view without sacrificing any of the screen

[AskMcGowan]

To explain this I am going to use a living room as an example.

The way it currently is set:

Put the camera in the corner of the room and catch the entire room - Fantastic!

Anyone walks into the view, it records - Fantastic!

Now say there is a glass on a table in the middle of the room. 

I can "crop" or "zoom in" on the table and the camera will shoot only that, so if someone touches the glass it records, but it only shows the table.

What I want: Show the entire room - BUT - only put the motion detection on the table, so the entire room is in the camera view but the recording only happens if someone is in the table view. This way I can see where the person went, who the person was, etc instead of potentially just the hand grabbing the glass.

[UK_Wildcats]

I understand the limitations of the PIR sensor and hardware.  I assume that the local base has to do some level of processing (even if minimial) for rules.   I am just wondering if there is a different way to perform the recording detection without using the PIR.   

To Mark's point, why can't we use constantly monitoring algorithms - i.e. using video comparative algorithms on the base side to determine if there is enough of a change.  I assume that the camera would have to setup to live stream and have a good enough bandwidth connection to base.  Because the algorithm would only be looking for a large enough variance between the video, the video resolution could be more lower.  In fact a lower resolution may help the algorithm effectiveness.  This would also help minimize bandwidth issues and reduce the amount of processing for the algorithm. Once it needed to record, the resolution could be increased.

If using video comparative algorithms, activity zones could be easily implemented.

Because of the battery consumption from live views, I assume that the camera to be constantly plugged in.  

Can the base actually handle the processing to do this?  What are the local base limitations?  Could a higher processing base that could perform this be in development?

[jasongabler]

@UK_Wildcats,

I believe what you suggest (and more) is provided by the new Arlo Smart service. 

jyg

[Pentarus]

Can someone tell me if the Arlo Smart plans (activity zones, person recognitions, etc) is available in Canada? I did Subscribe to the Premiere plan but I don't see these options anywhere in the settings.

[JamesC]

This feature is now available through Arlo Smart. Take a look here for more information on Arlo Smart and setting up Activity Zones: Introducing Arlo Smart - FAQ & Resources

[JamesC]

Pentarus,

Currently, Arlo Smart is only available in the US. An announcement will be made when it becomes available in other countries.

JamesC

[Pentarus]

Thanks JamesC,

That's a shame. I don't suppose anyone can give us an ETA for when this will be available in Canada.

[amirul]

I added activity zones in AP2, and it is plugged in, but when someone come in this predefined zone the camera does not trigger recording at all.  Any idea

[JamesC]

amirul,

Consider rebooting your base station and test again, do you still experience the same behavior?

JamesC

[UK_Wildcats]

The front of my house is similar to the letter "L"; therefore, I have two Arlo Pro cameras that face the street.   I have rules so that each camera records when either sees motion.  The garage  camera is much close to the screen, next to a tree and is constantly triggering when a car drives by or wind blows branches.  I have tried all the sensitivity settings.  To avoid constant recording, I basically have to the sensitivity all the way down which does not pick up people.   From what I read about Arlo Smart, it  seems to be more cloud based.  I want to setup zones to avoid recording and saving battery life.

[graycat]

Is it possible to set up multiple zones, so that I can have a outside zone and a bedroom zone and living area zone.

Can anyone explain to me how to do this.