I have implemented Green-Screen(Background Removal) functionality in my code using Kinect V2 sensor. Now i need to detect position of face points like eyes, nose and mouth.
Green-Screen is in the depth stream so using FacePointsInColorSpace method I'm not able to get proper position of the face objects.
Also one of the FacePointsInInfraredSpace method is available for that and it also tracks the objects but display wrong position. It's difficult to understand that Infrared & Depth steam both have 512 * 424 resolution so why it not detects the proper position of face points.
I have placed my IR stream face points detection image using FacePointsInInfraredSpace method at here : 
So how to get proper position of an face points in infrared stream so that i can use it inside the Green-Screen implementation.
Also, one of my another question is that How to get proper position of face-points when we uses image frame size manually Eg. In 1920 * 1080 resolution it gives proper image frames tracking for face-points but now if i set frames size is as 960 * 540 or any one which i want to set then how can i get proper position of an face-points.
