Amcrest Forum

So in working on a project involving the face detection API with the IP8M-TD2685EW-AI camera, I noticed in the API documentation that supposedly the event is fired off with actual face characteristics in the JSON array. However, in practice, the array I get has none of that data.

This is what the docs show:




However, in reality, this is what I get:




My goal here is to try and filter out repeat face detections of the same person within a short amount of time. For example, a person walks by the camera, lingers in front of it, shifting around, and each time their face turns towards the camera, it will fire off a face detection event. Because of the extra processing that takes place each time a detection event is received, it would be a lot more economical to filter out excess events of this kind. I could do that if I had those face characteristics, but not without them.

Any thoughts?

While i have that camera, I hate to say that is not one of the features for the camera model. You will need to buy a higher grade camera that has that feature built in. If you look at the picture below from one of my Dahua cameras. You can see where it is marked. This area is where I can pick what I want to show like Mack or Beard. Age Gender and so on.. However nice the 2685 is. That isn't a feature of the camera..