Non-rigid global alignment using thin-plate splines

A key challenge in reconstructing high-quality 3D scans is registration of data from different viewpoints. We present an algorithm for obtaining a globally-optimal alignment of multiple overlapping datasets in the presence of low-frequency non-rigid deformations, such as those caused by device nonlinearities or calibration error. The process obtains sparse correspondences between views using a locally-weighted stability-guaranteeing variant of ICP. Global positions for feature points are found using a relaxation method, and the scans are warped to their final positions using thin-plate splines.