How does Kinect calculate depth?

Question

I'm a little confused here.

How Kinect calculates depth : I understand that

• The IR projector produces a pattern that is reflected back and read by the IR camera.
• Now the IR camera knows the pattern for a certain depth. The difference between the incoming and known pattern is used to calculate the depth known by triangulation (using the proportionality of similar triangles).

Question 1 : Does the distance between the IR projector and the IR camera consider? I think not, because they are too close to consideration.

Question 2 : now we get the depth directly from the template. When do we use disparity mapto calculate depth?

Answer 1

The map of differences is basically the difference between the known and observed pattern, which you mention at the beginning. You use this when calculating depth.

The distance between the projector and the camera is also taken into account.

See the following figure:

Pr Zr Kinect Zo (, ). D - , d - 2D . f - , b - C L.

, , :

, , - Kinect" . . .