The JARtool project was a pioneering effort to develop an automatic system for cataloging small volcanoes in the large set of Venus images returned by the Magellan spacecraft. This package contains a variety of data to enable researchers to evaluate algorithms over the same images as used for the JARtool experiments reported in [Burl98].
Michael C. Burl MS 126-347, JPL 4800 Oak Grove Drive Pasadena, CA 91109 (818) 393-5345 Michael.C.Burl@jpl.nasa.govDate Donated: July 9, 1999
In addition to the images, there are "ground truth" files that specify the locations of volcanoes within the images. The quotes around "ground truth" are intended as a reminder that there is no absolute ground truth for this data set. No one has been to Venus and the image quality does not permit 100%, unambiguous identification of the volcanoes, even by human experts. There are labels that provide some measure of subjective uncertainty (1 = definitely a volcano, 2 = probably, 3 = possibly, 4 = only a pit is visible). See reference [Smyth95] for more information on the labeling uncertainty problem.
There are also files that specify the exact set of experiments using in the published evaluations of the JARtool system.
The image files are in a format called VIEW. This format consists of two files, a binary file with extension .sdt (the image data) and an ascii file with extension .spr (header information). There is a MATLAB utility function included in the data package that can be used to read the data. If you want to use something other than Matlab, you are on your own, but the format is fairly simple and can be understood by looking at the Matlab code.
The labeling files are provided in two forms. The .lxyr files are simple space-separated ascii containing label, x-location of center, y-location of center, and radius.
The Magellan Project Home Page
G.H. Pettengill, P.G. Ford, W.T.K. Johnson, R.K. Raney, L.A. Soderblom, "Magellan: Radar Performance and Data Products", Science, 252:260-265 (1991).
R.S. Saunders, A.J. Spear, P.C. Allin, R.S. Austin, A.L. Berman, R.C. Chandlee, J. Clark, A.V. Decharon, E.M. Dejong, "Magellan Mission Summary", J. of Geophysical Research Planets, 97(E8):13067-13090, (1992).
M.C. Burl, L. Asker, P. Smyth, U. Fayyad, P. Perona, L. Crumpler, and J. Aubele, "Learning to Recognize Volcanoes on Venus", Machine Learning, (March 1998).
P. Smyth, M.C. Burl, U.M. Fayyad, and P. Perona, Chapter: "Knowledge Discovery in Large Image Databases: Dealing with Uncertainties in Ground Truth", In Advances in Knowledge Discovery and Data Mining, AAAI/MIT Press, Menlo Park, CA, (1995).
Anyone seeking to publish results on this data should perform at a minimum the FULL SUITE of experiments defined in the Experiments_Images_Table included in the dataset and compare performance to that of the baseline JARtool system [Burl98]. Results on a small subset of images (e.g., HOM4) are not of interest and SHOULD NOT BE PUBLISHED. Send e-mail to jartool@aig.jpl.nasa.gov if you believe you have a compelling reason to deviate from this policy. We would also appreciate it if you would report your results (whether POSITIVE OR NEGATIVE) along with a brief description of your algorithm to jartool@aig.jpl.nasa.gov.
Assembly of this dataset has been carried out in part by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeroenautics and Space Administration.
Refer to the README file included with the dataset for more complete information.
The complete Magellan imageset is available in a large (~150) CD-ROM collection. Consult the NASA CD-ROM catalog for additional information.