-
Notifications
You must be signed in to change notification settings - Fork 2
Connecting to the Kinect
Brad Saund edited this page Jun 25, 2019
·
2 revisions
Victor's "eyes" are a Kinect sensor, publishing RGB-D images from the perspective of Victor's head.
Compressed images enter through a custom image listener and depth image listener.
Images are decompressed using OpenCV for C#. A custom "material" shader combines these images into a point cloud. Note this is computationally intensive. Using a 2080Ti this can run at approximately 10Hz using HD images.
The location of Victor's head enters as a pose message, generated from a Motion Capture (Vicon) system.