+ View the NASA Portal

Search OLBIN

Optical Long Baseline Interferometry News

2006 Interferometry Imaging Beauty Contest

WG Home | 2001 | 2002 | 2003 | 2004 | 2006 | Notes

2006 Contest Data
Four data sets are to be used for a single reconstruction.
The data represent different observations of the same target.
The baselines have been reconfigured between observations.
The files are as follows:
2006-03-03.fits
2006-03-04.fits
2006-03-05.fits
2006-03-06.fits
clue.pdf
clue.10.fits

2006 Test Binary:
The following is an additional sanity check.
The model is for 10 mas sep., PA=30, Dm=1, theta1=3 mas, theta2=1 mas.
double.fits

Entries are due Wednesday, 11:59 pm (Pacific Daylight Savings), 12 April 2006.

The 2006 IAU Interferometry Imaging Beauty Contest will be presented in Orlando, Florida, USA, in conjunction with the SPIE International Symposium on Astronomical Telescopes (24-31 May 2006). The IAU Working Group on Optical/IR Interferometry will meet at the IAU General Assembly in Prague, in August 2006.

The contest is being chaired by Peter Lawson (Jet Propulsion Laboratory) with participation from the following:

Organizers

Peter Lawson	Jet Propulsion Laboratory	Chair
William Cotton	National Radio Astronomy Obs	Principal Referee
Christian Hummel	European Southern Observatory	Generator of Contest Data

Tentative Dates

September - November 2005: Decide on rules and data sets.
14 October 2005. Deadline for Expressions of Interest.
4 November 2005. Rules and Regulations Decided.
7 November 2005. Submission of Abstract.
3 February 2006. All preparatory tests completed.
8 March 2006. Contest data set(s) released.
12 April 2006. Final images submitted to contest.
24 April 2006. Manuscript Submitted.
24-31 May 2006. Conference Date.

Participants

Name	Institution	Software	Contact Information
Fabien Baron John Young	University of Cambridge	BSMEM	baronmrao.cam.ac.uk jsy1001mrao.cam.ac.uk
Douglas Hope	University of New Mexico		dhope214gmail.com
Michael Ireland John Monnier	California Institute of Technology University of Michigan		mirelandgps.caltech.edu monnierumich.edu
Stefan Kraus Karl-Heinz Hofmann Gerd Weigelt	Max-Planck-Institute for Radio Astronomy		skrausmpifr-bonn.mpg.de khhmpifr-bonn.mpg.de weigeltmpifr-bonn.mpg.de
Serge Meimon Laurent Mugnier Guy le Besnerais	ONERA	WISARD	sergemeimonyahoo.fr Laurent.Mugnieronera.fr Guy.Le_Besneraisonera.fr
Joerg-Uwe Pott	European Southern Observatory		jpotteso.org
Sridharan Rengaswamy	Space Telescope Science Institute		sridharstsci.edu
Eric Thi�baut	Observatoire de Lyon	MIRA	thiebautobs.univ-lyon1.fr
Bob Zavala	USNO Flagstaff		Asotozavaol.com

Contest Objectives

Objectives of the 2006 contest may include

An understanding of how the entries cope with for example
1. different sorts of data sparseness
2. different noise levels
3. different source morphologies
Understanding the effect of any parameters/optional features supported by the entries.

Items [a], [b] and [c] should all be realistic for the interferometers available now and within (perhaps) the next couple of years, and the sort of science people anticipate being able to do. Understanding the above would all help with

observation planning
choosing a package to map your data
using the package optimally
knowing what features to believe in the map

Test Data

Christian Hummel produced two files (attached) in the OI-FITS format of simulated binary data with a simulated 6-station array, 15 baselines, and 10 triples.

Test Data Set 1: BSC1948.fits. Can be modelled by a binary with sep. rho=21.2 mas and pa=341.6 deg, flux ratio 5.75, component diameters of 0.6 mas. This information is being made available so you can check if you are able to fit the correct model parameters.
Test Data Set 2: FKV1137.fits. Also a binary, but more complicated.

Previous Contest Data

Previous test data sets are available on line, as well as the contest data from the 2004 contest. There are two data sets, as follows. Please note that the target names cited in the oi-fits files (Mizar and Altair) bear no relation to the data itself.

Contest Data Set 1: data1.oifits
Contest Data Set 2: data2.oifits

2006 Contest Data

Four data sets are to be used for a single reconstruction. The data represent different observations of the same target. The baselines have been reconfigured between observations. The files are as follows:

Observation data set 1: 2006-03-03.fits
Observation data set 2: 2006-03-04.fits
Observation data set 3: 2006-03-05.fits
Observation data set 4: 2006-03-06.fits
clue.pdf
clue.10.fits

The model image used to generate the contest data is not convolved with a PSF, but prior to comparison, the model and all submitted images will be convolved to a common resolution using a Gaussian PSF. The clue provided is a low resolution image to give the relevant FOV and has been deliberately modified such that its Fourier transform will not produce usable information. The FITS file header does contain the cell spacing and the plot displays the full field. Only information in the OIFITS files and knowledge of the field of view should be used in generating submitted images.

2006 Test Binary:
The following is an additional sanity check. The model is for 10 mas sep., PA=30, Dm=1, theta1=3 mas, theta2=1 mas.
double.fits

Contest Entry Checklist

Submissions to the contest will consist of simple FITS images with the following descriptive information provided either in the FITS image header or separately:

Pixel separation and orientation as well as any rotation and/or skew.
Pixel brightness units: e.g. Flux density per pixel or flux density per resolution element and in the latter case the equivalent number of pixels per resolution element.
Description of resolution in image including any convolution.
Provide cell spacing either in FITS header or separately.
Discussion of which features they consider real.
A visual representation of the image (contour, grayscale, etc.) may also be provided but the entry will be judged on the FITS image.
Some text would be greatly appreciated that would briefly describe your software and its strong points. This text will be edited and included in the contest paper.

Judging steps

William Cotton, the contest judge, will take the entries and perform the following manupilations as part of evaluating the entries:

Convolve to a common resolution.
Regrid onto a common grid. Since there is no astrometric information in the data this will require aligning on a feature in the image.
Difference from model and sigma/peak used as quality measure. This would be used if one or more entrants get basically the correct image, otherwise it would be which one looks closest.

There are also potential problem areas which may require modification to the evaluation.

Image alignment. Since astrometry is not part of the test, we'll have to align the images with the "answer" before comparison. Last year's test images had well enough defined features that they could be used. Depending on this year's test cases, something more complicated (cross correlation?) may be needed.
Normalization. We've tried to avoid this problem in the requirements for the image units but this may not be sufficient.

Contest Results

PowerPoint and PDF of awards presentation at SPIE in Orlando, Florida

References

Potential contestants are encouraged to consult the links at the top of this page, as well as the following two publications.

An Interferometry Imaging Beauty Contest,
P.R. Lawson, W.D. Cotton, C.H. Hummel, J.D. Monnier, M. Zhao, J.S. Young, H. Thorsteinsson, S.C. Meimon, L. Mugnier, Guy Le Besnerais, E. Thiebaut, P.G. Tuthill,
Proc. SPIE 5491, New Frontiers in Stellar Interferometry, 886-899 (2004).
A data exchange standard for optical (visible/IR) interferometry
T.A. Pauls, J.S. Young, W.D. Cotton, and J.D. Monnier
Pub. Astron. Soc. Pac. accepted (for Nov 2005 issue).

Peter Lawson

Updated: 4 April 2006

Maintained by Peter Lawson
MS 301-451, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena, CA 91109

Webmaster: Peter R. Lawson