0
votes

Im currently tracking a Skeleton Joint real world coordinate in multiple Kinects.

But these coordinates are expressed in relation to each Kinect coordinate space.

What i want to do is elect one of the Kinects as the "Master" and have all others transform their coordinates to this Master Kinect's coordinate space. I made a paint image to illustrate:

enter image description here

From what i ve seen on the internet this involves heavy math. Is there any ready code/library with functions that do this? Or even matlab code? Im using C# and Kinect SDK, but this doest not limits the language of an possible solution.

1

1 Answers

1
votes

Surely you can just use transformation matrices to achieve this? If you know the rotation and translation necessary to get from "Slave" to "Master" Kinect, you can just apply it to your point cloud detected by the "Slave Kinect".