Start a new topic

place 3d object [wt3] in android without image tracking and without location


1)I want to place the 3d model  [wt3] or any 3d model in android on plane surface i don't want image tracking and location for 3d model with your sdk . we want to buy your sdk but this is our business requirement. we want to add the 3d model and we can drag,drop,scale.

or can we use other android native 3d engine with opengl for achieving this.

2)and second thing is create problem with google panoramaView.

Thanks and Regards 

Swapnil  pol.

Hello Swapnil,

Regarding your first question, Instant Tracking would be the feature that is ideal for your use case. The best way to start is to download our sample and test Instant Tracking to see if this covers your needs. You can also refer to our documentation here. and you can check out this article as a basic first tutorial with Instant Tracking.




Thanks for reply..
We want 3D model/object which can move with camera.we don't want  object at particular location.
In instant tracking that object gets invisible if we move camera.

Hello Swapnil,

So if I understand your use case correctly, you wish to place an augmentation (in this case a 3D model) not in a specific location (like you can do with Instant Tracking) and without using a tracker (such as an Image or an Object) or GeoLocation. I am afraid this is not possible, since our technology and Augmented Reality in general requires either a marker, a GeoLocation or Instant Tracking features where the object is in a fixed position.

May I ask, if this is your use case (having a 3D model in your camera that follows you every time you move the camera) then why do you wish to use Augmented Reality?




Your question is right.
We have used Augment reality everywhere in app using Wikitude. But for one Use case We want to show that 3d model(wt3) as continuous view above wikitude camera.We are stuck on this last case.


what you want to do is possible with the JavaScript SDK, but you will need to use a (very simple) plugin and an AR.Positionable with it. The AR.Positionable object allows to specify the position of any AR.Drawable (e.g. AR.Model) and thus bypassing the tracking of the SDK itself. The intention of this feature is to allow users to implement their own tracking algorithms, but a fixed position can, obviously, be supplied as well.

We do have an extensive sample for this use case. You can find the corresponding documentation page here.

In the MarkerTrackingSample class, you would simply remove most of the code and add the following lines in the updatePositionables function:


wikitude::sdk::Matrix4 identity;

wikitude::sdk::Matrix4 trans;
trans[0] = 0.00962843f;
trans[1] = -0.765947f;
trans[2] = -0.288562f;
trans[3] = 0.0f;

trans[4] = 1.33069f;
trans[5] = -0.00890594f;
trans[6] = 0.0619125f;
trans[7] = 0.0f;

trans[8] = -0.0833194f;
trans[9] = -0.23075f;
trans[10] = 0.955458f;
trans[11] = 0.0f;

trans[12] = 0.0396224f;
trans[13] = 0.0450674f;
trans[14] = -0.396882f;
trans[15] = 1.0f;



If you use the car.wt3 model, it will float nicely in the middle of the screen using this transformation. For your use case and model the values of this transformation matrix will obviously be different.

Make sure to call enteredFieldOfVision once. That will set the positionable into the "currently being tracked" state. This causes it to be displayed until you call on exitedFieldOfVision.



You will have to handle handle some things yourself (device orientation change, providing a projection matrix, ...), but all these steps are demonstrated in the MarkerTracking sample and accompanying documentation.

- Daniel

Login or Signup to post a comment