Markerless Augmented Reality (AR) registration using the standard Homography matrix is unstable, and for image-based registration it has very low accuracy. In this paper,we present a new method to improve the stability and the accuracy of marker-less registration in AR. Based on the VisualSimultaneous Localization and Mapping (V-SLAM) framework,our method adds a three-dimensional dense cloud processingstep to the state-of-the-art ORB-SLAM in order to deal withmainly the point cloud fusion and the object recognition. Ouralgorithm for the object recognition process acts as a stabilizer toimprove the registration accuracy during the model to the scenetransformation process. This has been achieved by integrating theHough voting algorithm with the Iterative Closest Points(ICP)method. Our proposed AR framework also further increasesthe registration accuracy with the use of integrated cameraposes on the registration of virtual objects. Our experiments show that the proposed method not only accelerates the speed of camera tracking with a standard SLAM system, but also effectively identifies objects and improves the stability of marker-less augmented reality applications.
展开▼
