Simultaneous registration of multiple range views with markers (ICCAS 2007)

We propose a new model for simultaneous registration of multiple range views. The new model minimizes a combination of the tangent distance error metric and the sliding penalty which is defined by the marker correspondences. As a result, the proposed model is capable of reducing the sliding error in registration, especially for the object containing flat features. We use the kinematical geometry to linearize rigid body motions and approximate the registration model with a quadratic objective function, which leads to solve linear equations at each iteration. We also present the experiments which show that our registration algorithm is independent of the initial alignment and holds a global stability.