Start a new topic

Scene Picking sample in SDK7

Hi, In SDK7 there is a scene called "Scene Picking" I couldn't figure out what this scene does and I appreciate if you can help me.
I also wanted to ask if there are any ways to identify and differentiate places and locations to use different SLAM content.

Thanks


Hi,

If you tap 'Start tracking' instant tracking starts and build a point cloud in the background. If you now tap on the screen, the Wikitude SDK performs a check to see if there is some point cloud information at the screen location. This might fail if you tap on a area where no information is present e.g. a white wall or ceiling. But if there is some information in the scene at the screen location, you get a 3d coordinate at which you can position new augmentations or use as a reference.


What I don't understand from your last question is the sentence: `What I'm trying to do is to identify where the SLAM is performed and augment related content`.


Best regards,

Andreas


1 person likes this
Hello,

What this new example demonstrates is that basically you can scan and track a scene (just like with Instant Tracking) but then you can position 3D points in the scene you are tracking. In order for this example to perform better, the scene must be an environment with different objects and you try to place the 3D points on those objects. The further away on your scene you are placing the 3D points the smaller those points are.

Regarding your second question, with Instant Tracking it is not possible but maybe you could give it a try with Object Recognition. If you can share more details about your use case maybe I can direct you further.

Thanks
Eva

 


1 person likes this

Sounds like what m0j1 42420 is asking is if it is possible similar to what Tango did in that you can scan the area and save your content in their place and then come back to that location and have all the content in the same place where you left off. 

In Tango you could "learn" an area, place all your 3d content, save their positions and then load them when that area is detected.


1 person likes this

Hi Valerio,

Watch closely for our upcoming release. It might contain exactly what you're looking for ;)


Best regards,

Andreas

Thanks very much, I still couldn't get it to work. There's "start initialization" and "start tracking" and both just shows the grid. 
What I'm trying to do is to identify where the SLAM is performed and augment related content. Like this :
https://twitter.com/AndrewProjDent/status/888380207962443777

There are different scenarios like Indoor navigation,etc.

Thanks again very much

Hi, 


I'd like to use a feature similar Tango "scan the area and save your content in their place and then come back to that location and have all the content in the same place where you left off", is it possibile with wikitude? Thanks

Login or Signup to post a comment