Return to Research

Joint Registration of Multiple Point Sets

A Generative Model for the Joint Registration of Multiple Point Sets

European Conference on Computer Vision (Computer Vision – ECCV 2014)

An extended version submitted to IEEE TPAMI is available on arXiv:

Lecture Notes in Computer Science Volume 8695, 2014, pp 109-122
G. Evangelidis, D. Kounades-Bastian, R. Horaud, E. Psarakis



This paper describes a probabilistic generative model and its associated algorithm to jointly register multiple point sets. The vast majority of state-of-the-art registration techniques select one of the sets as the “model” and perform pairwise alignments between the other sets and this set. The main drawback of this mode of operation is that there is no guarantee that the model-set is free of noise and outliers, which contaminates the estimation of the registration parameters. Unlike previous work, the proposed method treats all the point sets on an equal footing: they are realizations of a Gaussian mixture (GMM) and the registration is cast into a clustering problem. We formally derive an EM algorithm that estimates both the GMM parameters and the rotations and translations that map each individual set onto the “central” model. The mixture means play the role of the registered set of points while the variances provide rich information about the quality of the registration. We thoroughly validate the proposed method with challenging datasets, we compare it with several state-of-the-art methods, and we show its potential for fusing real depth data.

Keywords: point set registration, joint registration, expectation maximization, Gaussian mixture model


Download PDF

Download Poster

A generative model for the joint registration of multiple point sets
G.D. Evangelidis, D. Kounades-Bastian, R. Horaud, E.Z. Psarakis,
European Conference on Computer Vision (ECCV), Zurich, 2014


author = {Evangelidis, G.D. and Kounades-Bastian, D. and Horaud, R. and Psarakis E.Z.},
title = {A generative model for the joint registration of multiple point sets},
booktitle = {European Conference on Computer Vision (ECCV)},
year = {2014},

Code & Data-set

We provide a Matlab code that implements the JRMPC algorithm (Joint Registration of Multiple Point Clouds) as presented in the above paper. We also provide a data-set, referred to as EXBI data-set, of 10 real point-sets captured when moving a TOF camera around a static scene (see the figure below). While an RGB value is assigned to each point owing to a rigidly attached color camera (hence the inaccuracies near discontinuities), the color information is used for visualizations purposes only.  Please see the README file and the Matlab demos.

EXBI data-set

Download Code (ver 0.9.4) and Data-set
(last software update: May 12th, 2015)
last update: May 24th, 2014 (a bug in variance initialization has been fixed, README file is updated)


Evolution of GMM means when running JRMPC algorithm Registration of EXBI dataset (10 point sets) with several algorithms (see the paper)

If you cannot see the videos, visit the following youtube links: [Video1] [Video2]