Multiple 3D Object Tracking using ROI and Double Filtering for Augmented Reality

We present a new method to track multiple 3D object that is robust and real-time processing. Also, fast object registration is available. Many augmented applications need to track 3D object using natural features in real-time, more accuracy, and registering or learning object immediately. But existing algorithm uses FERN for feature extraction algorithm that takes several minutes to register target object, because of its high accuracy performance. Our method provides not only fast registering time about 0.3ms but also high accuracy. This features accomplished by SURF for feature extraction algorithm, ROI and double robust filtering. Using our methods, multiple 3D object tracking with fast and high accuracy is available.