It needs to have either Canvas supported live feed available or a built in event trigger on facial recognition that posts a image. Looking to build a welcome and tour/endpoint based system with voice support in my house and need cameras that works for that.
Edit, forgot to mention, i´m allergic to subscriptions so the cameras API needs to be open or one time fee.
I’m not 100% sure what you’re looking for or how much smartthings integration you need, but the two systems that offer affordable facial recognition are Netatmo and Nest. (Expect more after iOS 14 launch is with HomeKit support for local facial recognition, which is a very impressive feature.) Also, ring recently filed a patent on a form of facial recognition, but that could mean two years out given their past timelines.
Of these two, Netatmo does not require a subscription and has better smartthings integration, although to be honest, not that much integration.
Netatmo also has an IFTTT channel for the camera, so you can get a little smartthings integration that way as well.
Sorry maybe should have been clearer, i dont necessarily need smartthings integration at this point. just a camera with a availible API for getting the video stream or facial recognition built in with an API. Dont think Nest (or Smartthings camera) camera offers any form of open API sadly or it would have been a great camera
Let me walk you through what im building and maybe it makes sense,
Camera detection, when a visitor opens the door Camera 1 is going to spot a face.
At this point it would either match a known face or a new face,
a) Known match - API call will be made to Googles Text to Voice API welcoming the person by name (have this working already based on static pictures). Running locally on the tablet next to them using the built in speaker.
b) New face – Text to Voice API asking the visitor if they want to register at the tablet to the left. This will also call the Fully Kiosk API to open up a visitor registration form to fill in name, prefilled with photo and estimated age etc. from Face JS API
When registered, a introduction to the tablet will be done with Text to Voice API.
And option to play endpoint information by saving a state of that Person and what endpoints has been plaid, see below
Camera number 2, 3, 4 etc. depending on the location will each recognize if that person has been through each camera/endpoint. If not, a simple Text to Voice will play unique to that endpoint.
It’s a rather simple application and logic, most of the APIs I need I’m already familiar with and the database structure is already done, just need to find a fitting camera as everything will need to be custom code written from scratch. see there are some companies that does complex systems för Museums etc. but they are 100k+
So as you can see its not really needed to have a ST integration altough it would be a plus, just a good camera with API.