Geek Answers Handbook

Real-time custom image marker detection using OpenCV on iOS

I would like some tips, perhaps more, about finding a custom image marker in real-time video. I use OpenCV, iPhone and the channel for the camera. By means of a custom image marker, I mean a predefined image, but it can be any kind of image (and not a specific marker). For example, this may be an image of some skyscrapers.

I have already worked with ARTags and understand how they are detected, but how can I identify this custom image and especially find out its position and orientation?

What makes successful detection of a good custom image?

thanks

+4
source share
2 answers

This answer is no longer valid since Vuforia is now a paid engine.

I think you should try Vuforia. This is an AR engine that can use any image you want as a marker. What makes a good marker for Vuforia high-frequency images.

http://www.qualcomm.com/solutions/augmented-reality

Vuforia is a free engine.

0
source

The most popular markers used in AR are

  • AR markers (simple form of QR codes) - those that are detected by AR takeit and others
  • QR codes. There are many examples of how to create / detect / read a QR.
  • Dot grids. Like the chess nets used in calibration. It seems that their detection can be more reliable than the classic chess grid. OpenCV has codes related to the detection of a grid of points in the calibration part. In addition, the OpenCV code base is a good starting point for retrieving 3D position and orientation.
  • Chess nets. Like dot grids. They were a standard calibration scheme, and some people used them to detect markers for a long time. But recently, they have lost their position for dot grids when some people have found that points can be detected with better accuracy.

Note

The grids are symmetrical. I bet you already know that. But this means that you cannot extract full orientation data from them. You will get a plane where the grid is, but nothing more.

Final note :

The code and examples for the first two are easy to find on the Internet. Many consider them the best. If you decide to use grid templates, you should enjoy the work of processing mathematics and images :) And it will take more.

+1
source

More articles:


All Articles