-
Notifications
You must be signed in to change notification settings - Fork 274
Marker detection Installation Guide
The basic steps on how to setup the projects in order to integrate marker detection in a DroidAR project
- [http://www.crystax.net/android/ndk-r4.php Crystax NDK r4] (Other versions might work on Linux, on windows this seems to be the only working NDK for this project.)
- The DroidAR Project
- [http://www.cygwin.com/ Cygwin] (Download the Setup [http://cygwin.com/setup.exe here])
-
Get the project files from http://droidar.googlecode.com/svn/trunk/ARMarker
-
Get the OpenCV project source code from https://github.com/billmccord/OpenCV-Android.
-
Paste source folders cv and cxcore into the jni folder of the ARMarker project. (The other files are not needed. If you want to use other features from OpenCV you might need to edit the Android.mk file in the jni and add other files too.)
-
Load the external tool configuration "Build AR.launch" into eclipse and edit the environment variables to match your cygwin root folder and your NDK root folder. The cygwin path is a windows path and the ndkdir is a unix path. Keep this in mind when using spaces.
-
Compile the library using the external tool configuration. After the execution there should be a libs folder in your project with a libopencv.so file. This can take longer than 5 minutes, so you might want to start doing something else while the code is compiling.
-
Set the DroidAR project as a library project for the ARMarker project.
-
Create your own project which has both projects as library projects. For an example of such a project see the MarkerDemos project in the repository.