A cross-platform Augmented Reality library written in C++11.

Strongly WIP, so don't expect to use this for anything real at the moment. :)

• Supported platforms:
  • Linux
  • Windows
  • Android
  • iOS
  • Unity3D-plugin (Windows, Linux and Android). Unity 5.3 required.
• Extendable pipeline architechture (see Architechture)
• Marker tracking: multiple markers
• OpenGL-rendering: only simple 3D-shapes supported for the moment
• Support for sensor data:
  • Android: gyroscope, rotation vector (virtual sensor)

• Support for more trackers
  • SLAM
  • Edge-based
  • Face
• Platform support for OSX (native and Unity3D-plugin)
• Platform support for iOS-Unity3D-plugin
• Improved marker detection / different marker data-fields
• Support for a proper rendering-library (Ogre3D or OpenSceneGraph)

Yarrar is modeled as a pipeline, which consists of four stages. Each stage has one or more stage-handlers, which process the data from previous stage and pass it forward to next stage. Pipeline configuration is done with JSON-based configuration file (see Pipeline Configuration for details). Pipeline stages are:

1. Dataprovider: Sensor data handling (ie. RGB-camera, depth-camera, accelerometer, gyroscope)
2. Tracker: Pose tracking (ie. using markers, SLAM)
3. SensorFusion: Pose accuracy improvement
4. Render: Rendering the 3D-models on top of camera image

Limitations in current implementation:

• Only one dataprovider with RGB_CAMERA_FLAG set is supported
• SensorFusion-stage is WIP

Yarrar depends heavily on C++11-features, so a fairly recent compiler is needed.

Library dependencies per compilation target:

* Currently not needed (doesn't support sensor fusion), but will become a dependency in the future.

See Compiling for detailed instructions

Compile and run tests on Linux:

mkdir build
cd build
cmake ..
make
make test ARGS="-V"

1. Build AAR (see Compiling).
2. If you built it with build_aar.sh-script, .aar gets automatically copied to example/android/app/libs.
3. Open and build example/android with gradle or Android studio.

1. Build Yarrar.framework (see Compiling).
2. Copy Yarrar.framework to example/ios-directory.
3. Open and build example/ios/yarrarexample.xcodeproj with Xcode.

1. See Quickstart. It will also compile example/desktop_webcam-example by default.

1. Compile yarrar.unitypackage (see Compiling).
2. Open up example/unity in Unity Editor. Basic structure should be set up.

• When importing yarrar.unitypackage to Unity on Linux following error might appear:

Multiple plugins with the same name 'libyarrar' (found at 'Assets/yarrar/Plugins/x86_64/libyarrar.so' and 'Assets/yarrar/Plugins/Android/libyarrar.so'). That means one or more plugins are set to be compatible with Editor. Only one plugin at the time can be used by Editor.*

Fix this by deleting yarryar/Plugins/x86_64 if you wish to deploy to Android or yarryar/Plugins/Android if you wish to deploy to Linux.

• Deploying to Android might fail, if you don't set correct architecture for .so-library. This can be done by selecting libyarrar.so-file from Project-view and setting CPU to ARMv7 from Inspector-window.