Use of Morphological Operators and Pattern Recognition Techniques for Sorting the First Data [electronic resource].
- Washington, D.C : United States. Dept. of Energy. Office of the Assistant Secretary for Defense Programs, 1999. and Oak Ridge, Tenn. : Distributed by the Office of Scientific and Technical Information, U.S. Dept. of Energy.
- Physical Description:
- 5,300 Kilobytes pages : digital, PDF file
- Additional Creators:
- Lawrence Livermore National Laboratory, United States. Department of Energy. Office of the Assistant Secretary for Defense Programs, and United States. Department of Energy. Office of Scientific and Technical Information
- Restrictions on Access:
- Free-to-read Unrestricted online access
- The application of morphological and pattern recognition techniques to the sorting of FIRST data is presented. FIRST (Faint Images of the Radio Sky at Twenty Centimeters) is designed to produce the radio equivalent of the Palomar Observatory Sky Survey, using the National Radio Astronomical Observatory Very Large Array. A catalog of source positions and Gaussian-fit parameters for the sources is available. The goal of this part of the project is the automatic classification of types of sources, since it is expected that there will be on the order of one million sources in the final catalog. The data consists of radio maps, each approximately 1500 x 1100 pixels and containing on average about 30 sources per map. The resolution of these maps is 5 inches (approximately 3 pixels). The sources range in size from a few pixels to 70 or 80 pixels maximum diameter and have various morphologies. Of particular interest is the extraction of ''bent doubles'', core-jets, and gravitational lens candidates. These interesting types constitute on the order of one percent of the sources. The emphasis in this presentation is on image-based pattern recognition. We report on the use of morphological operators for source extraction and compare use of decision trees and neural networks in the classification of source morphologies and accounting for the finite resolution.
- Published through SciTech Connect., 11/04/1999., "ucrl-jc-136286", Signal and Imaging Sciences Workshop, Livermore, CA (US), 11/11/1999--11/12/1999., and Proctor, D.
- Funding Information:
View MARC record | catkey: 13824881