A new registration method for mixed augmented reality environments
This data was imported from Scopus:
Authors: Shi, X., Wan, T.R., Tang, W. and Zhu, Y.
Journal: Applied Mechanics and Materials
In augmented reality applications, one of the challenging tasks is to solve environment registering using only partial information of the targeted environments or physical features. We propose an improved method to solve the problem. Our vision based method estimates the relative camera pose between two calibrated views, combining marker points with environmental physical features to increase the effectiveness and the accuracy of the registration process. Experimental results show that our approach can recover the 3D structure information of actual scenes effectively, and the system developed satisfies the requirement of augmented reality (AR) applications successfully. © (2012) Trans Tech Publications, Switzerland.