Input Image Size

October 8, 2010 - 9:14am #1

[=&;;quot]The size of the target input image depends entirely on the application. In our recommended scenarios of near field usage environments the real Image Target will have a typical size of 10-40" (20-100 cm), although smaller and larger sizes are certainly possible. Note that we are referring here to the actual size of the printed trackable.[/]

Topic locked

Detecting Certain part of the 3d Object Using Object Scanner

June 11, 2019 - 10:22pm #1

Is it possible to detect certain parts of the physical object using Vuforia Object Scanner? Or is it possible to create Object Dataset File for a specific part of the object and try to detect it? 

I have a rectangular DVD box and I am trying to detect different parts of it.

Log in or register to post comments

Detecting 3D object with image target

June 11, 2019 - 4:47am #1


I am trying to detect different parts of the DVD box (rectangular box) using image target. For the flat surfaces, I am able to detect it. But as soon as I connect cables and try to detect, it's not detecting at all. 

Please note that both images, (with and without cable) are part of my image target database. 

Log in or register to post comments

Extracting Target Pictures from Dataset DAT File

June 9, 2019 - 9:06am #1

Hi Everyone,

I know that DAT file actually is a ZIP file including:

  • config.feat
  • Target#01.kpts
  • Target#01.tex.jpg
  • Target#02.kpts
  • Target#02.tex.jpg
  • ...

JPG file is a scaled down grayscale version of the original target image.

Log in or register to post comments

AR Apple Problem

April 10, 2019 - 10:48am #1

I've got a question...
I have an apple as an example, I want to show something on this apple.
The apple can be anywhere ...
How can I get something displayed on the apple?
I've already tried this:
Apple photographed, then deleted background of the apple picture.
Inserted as a picture and put a text in a picture, this on the apple picture should now be displayed.
Somehow the apple is not recognized because from the background nobody is black ...
Does one have an idea how I could do that?

Kind Regards

Log in or register to post comments

Object Terrain vuforia 2018

March 27, 2019 - 5:20am #1

Good morning everyone,

I have the following problem:

I created the camera = ARCamera

I created the image = imagetarget

I created the terrain in gameobject> 3d> terrain

I put the terrain inside imagetarget.

I created a 3d cube and put it in the imagetarget to appear on top of the terrain.

When the cell phone camera identifies the image of the vuforia appears the terrain and the cube, but it does not align with the image as the cube stays.

what should I do or what am I doing wrong?


Log in or register to post comments


March 26, 2019 - 8:26am #1

I'm going to make application that will be recognize car wheel and add new rims on the wheel, can you help , how I can make target with  car-wheel, app must recognize not only my private wheels

Log in or register to post comments

What is the maximum number of image targets that can be realistically simultaneously tacked?

March 3, 2019 - 2:16pm #1

What is the maximum number of image targets that can be realistically simultaneously tacked? At least, particularly, on an iPhone? I see very little information in the way of this. I understand that Vuforia has lifted its simultaneous tracking limit, but what is that number in practice?


I've tried tracking multiple cards at once, but have been, so far, unable to track more than 5 or 6. Would a VuMark perform better at tracking multiple images? Say, 11?

Log in or register to post comments

Managed to Detect All Targets but Only a Fixed Number of Videos is Playable

February 21, 2019 - 7:03pm #1

Hi there. I am currently developing a Android app that can perform augmented reality video playback for my wife-to-be. There are a total of 59 target images and 59 videos (originally 1080p; transcended to 270p via H265). Weirdly, after building the apk (approximately 700MB), I can detect all the image targets but can only play 16 of them. The rest only shows a play button which can't play the video when touched. I have attached a video ( for your reference.

Log in or register to post comments

Using QR codes instead of VuMarks

February 4, 2019 - 1:38am #1

My goal is to use qr codes instead of vumarks. I don't want to have to manually create hundreds of vumarks or image targets it needs to be scalable. I am using a 3rd party library (ZXing) to scan for qr codes and user defined targets. My solution works well for a single qr code but vuforia is unable to differentiate between different qr codes so problems arise when I want to scan and track multiple ones simultaneously. It seems the similarity threshold for image detection is too low as they are 5 star targets. If anyone has any suggestions it would be greatly appreciated.

Log in or register to post comments