Do I need to map skeletal data into color image point? - kinect/XNA - kinect

Question

I have been working on a project where I need to interact with 3D objects using a hand joint, now I do that using distance formula but however . . i'm using hand joint (X,Y,Z) and using a formula to convert them to 3D space to be able to have the interact with the Models.
Instead of using a formula to do the conversion . . do i actually need to use the function MapSkeletalpoint to color image point??
To give an idea about what i'm doing:
var rightHands = playerSkeleton.Joints[JointType.HandRight];
var rightHandsX = rightHands.Position.X;
var rightHandsY = rightHands.Position.Y;
var rightHandsZ = rightHands.Position.Z;
HandX = rightHands.Position.X;
HandY = rightHands.Position.Y;
HandZ = rightHands.Position.Z;
foreach (_3DModel s in Solar)
{
x_mod = (float)Math.Floor(((HandX * 0.5f)) * maxWidth);
y_mod = (float)Math.Floor(((HandY * -0.5f)) * maxHeight);
z_mod = (float)Math.Floor((HandZ) / 4 * 20000);
if (Math.Sqrt(Math.Pow(x_mod - s.modelPosition.X, 2) + Math.Pow(y_mod- s.modelPosition.Y, 2) + Math.Pow(z_mod - s.modelPosition.Z,2)) < 20)
{
sound.Play();
Console.WriteLine("1" + "handx:" + x + "," + " " + "modelPos.X:" + s.modelPosition.X + "," + " " + "handY:" + y + "modelPos.Y:" + s.modelPosition.Y);
}

Answer 1

Actually, it's not easy to create effective hit test especially when there is larger number of objects.
The thing you need is a collision system. My recommendation to you is that you should use some existing, for example BEPU might be helpful:
http://bepu.squarespace.com/
You should bind the position of an object with the position of the hand (reported by Kinect) and BEPU will do the rest (it will trigger a handler in case of collision).
Just a small advice - it's always good to use relative position of the hand (e.g. to the shoulder). Using absolute coordinates reported by Kinect is not always desired in Kinect-based games (but you know better what you want)