Log in or register to post comments

Preprocess camera image prior to augmentation?

April 4, 2013 - 2:33pm #1

Is there any way to add custom process to the image between the camera feed and the uforia augmentation?
Some sort of visual filter that could help improve performance, perhaps?

Preprocess camera image prior to augmentation?

April 5, 2013 - 9:51am #7

We actively track feature requests as they arrive and hold regular reviews to evaluate them for future integration to the SDK. So you don't need to worry  that yours is at the end of a long list ;)

Also please consider adding the cases you've described to your wish list post. It's always helpful to know how devs intend to utilize the features that they are requesting.

Preprocess camera image prior to augmentation?

April 5, 2013 - 9:20am #6

Something that comes in mind right ahead is the possibility to apply computational filters to reduce glares - it is very hard to recognize trackers made on glossy surface, but there are algorithms that help reduce glaring, or augmenting contrasts.

 

Also, tweaking light exposure or even programmatically giving the option to the user to adjust the light exposure of the camera could end up in better tracking depending on where you are. Or this could be controlled by the device's light sensor.

I find it is hard to track in lower light environment such as subway or cafés, and sometimes also in high brightness environment: outdoor in full sunshine, the camera doesn't necessarily react properly and some trackers are washed off.

This is even more the case when we are using less than 4-star trackers - I know in a perfect world we'd have controlled lighting and 5-star images, but reality doesn't always permit that and having control over the way data is fed to the recognition engine could improve results - for those who wants to use it.

 

I have posted in the feature request thread. This list is very long, are there any chance those features are considered anytime soon?

Preprocess camera image prior to augmentation?

April 5, 2013 - 9:04am #5

We've seen a few related requests, for access to luminance levels and exposure settings to support more accurate scene lighting for example.

What is the use case that you have in mind that will benefit from preprocessing the image?

Preprocess camera image prior to augmentation?

April 5, 2013 - 8:23am #4

Hi, I understand very well your request; beside the quality/speed settings tha you already mentioned, the API does not offer other specific adjustments of the image which is fed into the Vuforia engine;

but this is definitely a good candidate as a feature request that you could post in our wish List here:

https://developer.vuforia.com/forum/general-discussion/wish-list

Preprocess camera image prior to augmentation?

April 5, 2013 - 8:18am #3

Thanks Alessandro.

 

I already do processing to the camera image. What I would like is to be able to control what is fed to Vuforia. For instance, a specific application might have different need in high luminosity or in low luminosity, and applying specific filtering might be something we'd want to do.

Perhaps there's some settings in Vuforia (other than speed/quality) that permits such interaction in the preaugmented image?

 

Another thing I could see is controlling how the camera is behaving. Right now all I can see is the focus control, can we decide how the camera adjust to brightness? I often see the passage from brightness to darkness ending in a suboptimal aperture setting, and CameraDevice.Instance doesn't seem to help in that regard.

 

Regards,

Preprocess camera image prior to augmentation?

April 4, 2013 - 10:51pm #2

Hi, technically you could retrieve the camera image and do some processing on it, as explained in this tutorial:

https://developer.vuforia.com/resources/dev-guide/unity-camera-image-access

However, you won't be able to feed the processed image  back into Vuforia, so I don't see an obvious way of doing what you describe.

 

Log in or register to post comments