0

I've been working on something which involves using the Oculus rift and vuforia. let me explain...

The overall objective is a pretty simple concept, I would like to be able to render the image from the webcam into the oculus rift and be able to detect multi image targets and display 3d models. I have the camera feed going to the rift and its working fine but the 3d model is not being drawn to the correct perspective and it seems to be outside of the oculus vision circles.

I have attached few images which i hope will make it a bit more clear.

enter image description here

When I disable the Oculus and just use the webcam everything lines up perfect as expected.

I think it may have something to do with QCARBehaviour script attached to the target, however I am unable to get anywhere.

Any help would be appreciated.

Wooble
  • 87,717
  • 12
  • 108
  • 131
steveybrown
  • 293
  • 2
  • 11

1 Answers1

0

have a look at this post at the Vuforia forum. They also want to use two cameras. Therefore they modified the Background Texture Access sample for the correct video-streams. But that's not your problem.

I think what you need is a second camera in your scene, one for each eye:

  1. Disable the camera inside the ARcamera
  2. Add two camera to it (as Child)
  3. Set Clear Flags to Depth only and Right Camera x = 0.63 (right eye).
  4. Set the Normalized Viewport rect (X) for the left camera to -0.5 and the right camera to 0.5 (split the screen).

After that you have to change some settings on the cameras for the correct positioning of a detected object. In my case I changed the FOV of the cameras (try and error, works best with the Multi-Targets sample).

Look at the first Link for more information!

user1972807
  • 61
  • 1
  • 5