Log in or register to post comments

User-defined targets through Cloud Recogntion

May 23, 2013 - 6:59pm #1

Hello there!

 

I've been experimenting with Vuforia for my graduation project for a few months now and I'm quite impressed! The tracking works really well and the community is really eager to help. I'm using Unity3D because of the intuitive workflow

I have read quite some threads on the forum and asked a few questions already. Now I thought it was time to open a thread to describe the challenge that I have. It could very well be that some of these questions have already been asked.

Basically I want to combine the functionality of the User-Defined Targets sample app and the Cloud Recognition sample app. This is a description of the user experience:

 

- The user scans an image, the app uses Cloud Recognition to find the target.

- If the image is not recognized, so isn't an ImageTarget, the user can choose to add the Image, similar to the User-Defined Targets sample app. However, the image needs to be uploaded to the Cloud database instead. That way the other users can track this image as well and the targets will not be lost after closing the app. NalinS already told me this is possible by uploading the image to my own server, and uploading it to the Vuforia Cloud using VWS API. I also know that this process can take 15-30 minutes.

- After adding the target, the user can provide text, which is stored in an external database which other users can add as well. 

- If the image is recognized by Cloud Recognition, the ID of the target is sent to the server, which reads the texts of all the users. These texts are then used to create the augmentation layer

 

I have little experience with server-side programming, so I'm not really sure on where to begin with this project. A few questions that come to mind:

- Is there anything in this workflow that is not possible? Or am I perhaps forgetting important parts?

- Do you know of an app who use this workflow of letting users add ImageTargets to the Cloud database? Or are there perhaps tutorials or samples that I should see?

- Are there any useful, simple services to temporarily store the images and the database of augmentation texts? I've seen smartfoxserver, but thought that might be an overkill.

- The Java examples to access the VWS API look interesting. How can you run these on a server though?

 

I really hope you can help me with this project. I will add extra information along the way.

 

Ciao,

Robin

User-defined targets through Cloud Recogntion

August 13, 2014 - 2:54pm #4

I second that question. 

I need to be able to push User Generated Image targets up to a server. 

As of 2.8, I really expected to be able to do this, considering that I had access to ImageTargetData. Unfortunately, as of 3.0.9, structures related to the actual image data have been wrapped up in a *.dll. 

Does anyone have experience with this, is it still possible?

Bryan

User-defined targets through Cloud Recogntion

May 29, 2013 - 5:27am #3

Thanks AlessandroB

 

I'm aware of the fact that  my questions aren't directly related to Vuforia. I was hoping some users could point me in the right direction for uploading imagetargets to the cloud database.

User-defined targets through Cloud Recogntion

May 24, 2013 - 12:31am #2

Hi Robin, thanks for your comments.

Your question is very broad (and partially not related to Vuforia, e.g. server programming topic);

I'll just add a couple of answers/comments here; hopefully other Forum users might help on the server specific questions (smartfoxserver, etc.)

 

- is there anything in this workflow that is not possible? Or am I perhaps forgetting important parts?

The workflow you describe (uploading image to Cloud and then reco + send ID to another server + read text on server) seems logic and doable, i.e. you should be able to implement it; although it can be quite some work.

- Do you know of an app who use this workflow of letting users add ImageTargets to the Cloud database? Or are there perhaps tutorials or samples that I should see?

Not that I am aware of, but I presume many other developers may have implemented something similar.

 

 

 

Log in or register to post comments