Hello,

Ground Plane tracking performance can vary from device to device.

Ground Plane utilizes images from the device camera to understand the basic geometry of your environment. Therefore, it is necessary to stage Ground Plane experiences in settings that enable the camera to accurately capture the details of the surroundings. 

Some recommendations:

• Stable lighting conditions
• Moderate lighting - not too bright or too dark
• Avoid glare and dark shadows
• Surface details help to reveal geometry and improve accuracy and performance

We've found that in areas with low detail surfaces we were able to improve the experience by taking additional time to scan the environment before performing the first hit test. While not always guaranteed to work, we were able to do 6DoF/positional tracking in a wide variety of environments such as low detail carpet, tiles, and pavement.

Thanks,

Vuforia Engine Support

First thanks for your answer. If I understand, the problem is the quality of performance of the tablet. 

I don't find the scan environnement in your website, can you give me the URL to have the explanation to test it?

Thanks,

Hello,

"Scan the environment" is a phrase I used to describe a user moving the device camera around, viewing the space where they'll place the augmentation *before* performing a hit test.

The idea here is that if a UX element can be added that prompts the user to do this kind of a motion before they register virtual content to the environment, there is a greater chance for robust 6DoF Gound Plane tracking experience.

Thanks,

Vuforia Engine Support

I listen your explanation and i try an experience. (all in the same condition and respecting the condition you give me). I see 30 second the environnement with the tablet before immersion for both of them, no wall, only ground with caractéristics points for the detection :

-First, I create an immersion with a little car (20 cm), I will no have problem, insertion at the good point, good orientation, when i move around, very good stability.

-Second I test with a big car, real size (1m50-2m), I have some stranges reaction. Sometimes, the car don't have a good orientation, it takes a random parameter. Sometimes i have a little car or the car go in the sky or move when i turn around the model. 

So my questions are : Normally, the size of the model don't play with the stability of it? What could be a solution to improve it?

My thanks.

Zephyr64

I make a test to show somes troubles. 

For both experience, I had the same condition, good light, detection of the ground during 15-20second, no wall :

-First, I try to use ground plane with a little object (little car 15cm²). I had no problems and I could turn around with good stability.

-Second, I try this time with a real size car (1m50-2m). Sometimes, the importation wasn't good, there are a reduction of the size, and most of the time, the car move on the sky or move when i try to turn around. 

So my questions are : it's strange but the systeme depend on the size of the object? How can I improve this problem of stability? What could be the problem if it's not the car?

Thanks,

Zephyr

"Scan the environment" is a phrase I used to describe a user moving the device camera around, viewing the space where they'll place the augmentation *before* performing a hit test.

Can you please provide a best practice example of a "scan the environment" phase before performing hit test?