cancel
Showing results for 
Search instead for 
Did you mean: 

Hololens technology

benplace
Rising Star
Why doesn't the Hololens require a sensor camera?  I have googled this and seen a ton of technical information, but could someone explain, in laymans terms, how it works and if this tech might come in the future with CV2 or 3?
9 REPLIES 9

Chassit
Heroic Explorer
My guess is that it doesn't need to know where you are looking and how your head is turning. VR headsets needs to rely on sensor cameras because they need to track the position and angle of your head in almost real time, to adjust your view in the virtual world. With HoloLens, that is entirely obsolete. You are seeing the REAL world in real time all the way through. The computer doesn't need to track anything, just to paint some extra stuff over your real vision.

All HoloLens needs to do is to generate a spatial map of your environment, and recognize objects when they pop into your vision.

Edit: HoloLens has depth camera (akin to Kinect) built into it, telling the headset how far you are from objects in your field of view and determine the distance and direction of your movement based on this information. That might also help.

DelmarKane
Expert Protege
The way I understand it, it uses a scanner, very similar to the Vive base stations.  Only it is able to detect reflections and the angle of the reflections.  Not entirely sure how.

I suppose it could be optical detection but I remember one of the developers specifically saying it was an active scan, not a passive one.  That the active scan was always happening and that is how the Hololens was able to send messages to the robot about a person standing in its assigned path.

benplace
Rising Star
I have a hololens and I find it very cool to be able to walk all over my house looking for clues in a crime scene investigation game.  The thought in my head is wow, why cant my Rift do this?

FX2K
Heroic Explorer
I have no idea, but did find this via Google:


The HoloLens has two Kinect-like depth cameras on it (facing forward-left and forward-right), which continually scan the environment of the user. These cameras create three-dimensional geometric representations of objects in the environment....... etc etc etc


https://www.reddit.com/r/HoloLens/comments/3aekn3/eli5_how_the_hell_does_the_realtime_tracking_work/
CV1: Ordered 6th Jan 2016 - Est Delivery Some time in May... DK2: Ordered: 8th of Aug 2014 - Delivered: 14 Oct 2014

Dreamwriter
Rising Star
HoloLens includes a little Kinect (depth sensor), which scans the room you are in and tracks the furniture and walls (and your hands). Using that it can figure out how you move in 3D space, because it sees those walls and furniture getting closer or farther and stuff. Google has similar tech made for cell phones (Project Tango), I wouldn't be surprised if their Daydream VR uses it for inside-out motion tracking.

Chassit
Heroic Explorer

benplace said:

I have a hololens and I find it very cool to be able to walk all over my house looking for clues in a crime scene investigation game.  The thought in my head is wow, why cant my Rift do this?



I'm all for AR. It has a lot of things sorted out the right way, like:

1. Using your real vision as the canvas, so you don't need a damn powerful GPU to reinvent the wheel in a virtual world.

2. Determining its own position by your location relative to surrounding objects, making tracking sensors obsolete.

3. Because of this massively reduced computing workload and power consumption, it can be made into an untethered device.

4. Being able to see your real world while using it means almost zero safety concern.

Would have got a HoloLens if not for the $3000 price tag...

DelmarKane
Expert Protege
Thanks all.  I see it is dual optical sensors now.  Not sure why I thought it was a laser of some type.  Probably due to the video I saw or something.

Dreamwriter
Rising Star
I think it still might include a laser - that's how Kinect works. Kinect shines a grid of a million tiny IR dots around the room, and then uses a camera to see how those dots change in size and orientation to figure out the shape of things. And as an object gets closer to the camera, those dots get larger so it knows the object us getting closer. Think of it like constellation tracking, but amped up to the extreme.

benplace
Rising Star
Thanks everyone, this makes sense now to me.  I can see in the future this same technology with an Oculus HMD that sends screens and input commands to and from a PC wirelessly (Like Citrix) and uses a passthrough camera.
That would be good for a VR or AR experience and allow untethering.