Data registration in 3-D scanning systems
LI Ren-ju · Journal of Tsinghua University(Science and Technology) · 2004
A 3-D measuring data registration method for 3-D scanning system was developed based on feature points and iterative closet point (ICP) algorithm. The feature points are first identified and matched. Then, the rotation matrix R and the translation vector T are found using the singular value decomposition algorithm. Thus, initialized by the former result, the ICP algorithm then leads to a perfect registration after improving on finding closet distance and deciding when to terminate the iterations. The basic principle of this method is discussed, and a registration application example of 3-D measuring data taken from different viewpoints is carried out to show the perfect feasibility of this approach.