HuskyLens is an easy-to-use AI machine vision sensor. It is equipped with multiple functions, such as face recognition, object tracking, object recognition, line tracking, color recognition, and tag(QR code) recognition.
It seems like many of the projects posted online using HuskyLens's Object Recognition or Object Tracking modes tend to involve rigid, unchanging shapes. Would the AI for these two functions be able to recognize and track shapes which are more flexible or amorphous?

For example, would HuskyLens be able to recognize a human walking though the forest at some distance away, and then track this person? I'm assuming this would pose a big challenge, since the shape of the person being tracked would be constantly changing while being obscured by branches and leaves. There would also be quite a lot of additional motion in the scene from the foliage blowing in the wind.

Also, is it possible for HuskyLens to measure the distance to the recognized object? Or, would one have to source an ultrasonic or laser sensor for this? This may be an issue in a forested environment because of all the foliage which may get in the way of the target. I haven't really figured out this challenge yet...

I'm asking because I am doing some initial research before starting a fun little project, and I want to make sure HuskyLens is the right choice before I commit. My goal is to build an autonomous sentry turret for airsoft/paintball matches using arduino and (hopefully) a HuskyLens. I already have the pan-and-tilt functions figured out from previous projects.