Effective scaling registration approach by imposing emphasis on scale factor
This Letter proposes an effective approach for the scaling registration of point sets. By imposing the emphasis on the scale factor, the authors design a novel objective function, which can avoid the phenomenon that the scale factor converges to zero. Subsequently, the well-designed function is optimised by the proposed scaling iterative closest point algorithm, which can calculate the optimal scaling transformation by iterations. The proposed approach is validated on public data sets, demonstrating its superiority over previous approaches on efficiency and robustness.
