Robust Simultaneous 3D Registration via Rank Minimization

We present a robust and accurate 3D registration method for a dense sequence of depth images taken from unknown viewpoints. Our method simultaneously estimates multiple extrinsic parameters of the depth images to obtain a reg-istered full 3D model of the scanned scene. By arranging the depth measurements in a matrix form, we formulate the problem as a simultaneous estimation of multiple extrinsics and a low-rank matrix, which corresponds to the aligned depth images as well as a sparse error matrix. Unlike previ-ous approaches that use sequential or heuristic global reg-istration approaches, our solution method uses an advanced convex optimization technique for obtaining a robust solu-tion via rank minimization. To achieve accurate computa-tion, we develop a depth projection method that has mini-mum sensitivity to sampling by reading projected depth val-ues in the input depth images. We demonstrate the effective-ness of the proposed method through extensive experiments and compare it with previous standard techniques. 1