Need help with opencv-markerless-AR-Mobile?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

215 Stars 103 Forks 2 Commits 4 Opened issues


Markless AR transplanted to iOS and Android based on takmin's code.

Services available


Need anything else?

Contributors list

# 246,220
2 commits

The Markerless Augmented Reality by OpenCV on Mobile phones

based on takmin project:


Windows version.

Can be found here:

I have compiled the source code under windows 8. also updated the opencv version to 2.4.9. The visual-studio-2012 project can be found under the "src" folder.

this is my testing video (win8/vs2012):


Android version (not finished.)

I have transplanted the recognize part on the android platform with the opencv 2.4.9, I have to compile libnonfree.a from the opencv source.

This based on the opencv-android-mixed-demo, which has Java side for camera handling and also NDK side for native opencv operations. The recognize-part-src was compiled with NDK, then running under my Android Phone, but it's too slow and very hard to debug the native code, so I continue the work with iPhone 5C (much better than my android phone).


iOS version :


Building and running with opencv-2.4.10-ios-version under iphone 5c, I have made some changes to make it running quicker. Now the recognize part cost 50-80 ms, while the tracking part cost only 15-25 ms, much quicker. But the speed is still not so good with mobile phones. Also , when tracking , there are accumulated-errors which can cause the pose very easy to fail ( the pose calculated by opencv opticalflow can have some error each frame, and will be accumulated to the last frame).


BTW, I'm working on a Markless tracking project on my current job, video can be found:

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.