Live video with object detection

SOLVED
xamatgar
Conversationalist

Live video with object detection

Hello, 

 

Is there any way to get a link of the Meraki live video with the yellow boxes (with the objects detected)? Thank you in advance

1 ACCEPTED SOLUTION

You could use motion alerts, via Webhook, to trigger the capture of footage: 

https://documentation.meraki.com/MV/Advanced_Configuration/Motion_Alerts#Alert_Only_on_People_Events

https://developer.cisco.com/meraki/api-v1/#!get-device-camera-video-link

 

There's a beta tool available to trigger such alerts only on people detection too.   Right now I don't believe that extends to vehicles.    Bear in mind that if you also triggered the capture of screen grabs, you could send the output to AWS Rekognition (for example) which could detect vehicles, people, text and all sorts of other things.

 

You may also find MV sense particularly useful in pre-processing and adding more granular information:   https://meraki.cisco.com/lib/pdf/meraki_datasheet_mv_sense.pdf 

View solution in original post

5 REPLIES 5
GreenMan
Meraki Employee

No - object detection is performed post-capture.

xamatgar
Conversationalist

So, there is no API to extract the video with the detections even post-capture right?

You could use motion alerts, via Webhook, to trigger the capture of footage: 

https://documentation.meraki.com/MV/Advanced_Configuration/Motion_Alerts#Alert_Only_on_People_Events

https://developer.cisco.com/meraki/api-v1/#!get-device-camera-video-link

 

There's a beta tool available to trigger such alerts only on people detection too.   Right now I don't believe that extends to vehicles.    Bear in mind that if you also triggered the capture of screen grabs, you could send the output to AWS Rekognition (for example) which could detect vehicles, people, text and all sorts of other things.

 

You may also find MV sense particularly useful in pre-processing and adding more granular information:   https://meraki.cisco.com/lib/pdf/meraki_datasheet_mv_sense.pdf 

If you're looking to consume something existing, rather than 'build-your-own' - maybe explore what's already available in the Marketplace?    https://apps.meraki.io/apps/?meraki-product-integration=Meraki%20Cameras

PhilipDAth
Kind of a big deal

>So, there is no API to extract the video with the detections even post-capture right?

 

You can use MQTT and it sends you the rectangle coordinates for every object in the frame.  You can then draw the rectangle on top of the image yourself.

 

This is the specific MQTT reference:

https://developer.cisco.com/meraki/mv-sense/#!mqtt/raw-detections-a-list-containing-object-identifie... 

Get notified when there are additional replies to this discussion.
Welcome to the Meraki Community!
To start contributing, simply sign in with your Cisco account. If you don't yet have a Cisco account, you can sign up.