SKICAT: Sky Image Cataloging and Analysis Tool

In astronomy and space sciences, we current ly face a data g lut crisis. The problem of dealing w i t h the huge volume of data accumulated f rom a variety of sources, of correlating the data and extract ing and visualizing the impor tan t trends, is now fu l ly recognized. This problem wi l l become more acute very rapidly, w i th the advent of new telescopes, detectors, and space missions, w i th the data flux measured in terabytes. We face a cri t ical need for in format ion processing technology and methodology w i th which to manage this data avalanche in order to produce interesting scientific results quickly and efficiently. Developments in the fields of machine learning and AI can provide at least some solutions. Much of the future of scientific in format ion processing lies in the implementat ion of these methods. We present an appl icat ion of supervised classification to the automat ion of the tasks of cataloging and analyzing objects in digi t ized sky images. The Sky Image Cataloging and Analysis Tool ( S K I C A T ) was developed for use on the images result ing f rom the 2nd Palomar Observatory Sky Survey (POSS-II) conducted by the Cal i fornia Inst i tu te of Technology (Caltech). The photographic plates collected f rom the survey are digit ized at the Space Telescope Science Inst i tu te. Th is process w i l l result in about 3,000 digi ta l images of 23,040 x 23,040 16-bit pixels each, to ta l l ing over 3 terabytes of data. When complete, the survey w i l l cover the entire northern sky in three colors, detecting v i r tua l l y every sky object down to a B magnitude of 22. Th is is at least one magnitude fainter than previous comprable photographic surveys. We estimate that there are on the order of 10 galaxies 10 stellar objects ( including over 10 quasars) are detectable in this survey. Th is data set wi l l be the most comprehensive large-scale imaging survey produced to date and w i l l not be surpassed in scope unt i l the complet ion of a fu l ly d ig i ta l all-sky survey. The purpose of S K I C A T is to enable and maximize the extract ion of meaningful in format ion f rom such a large database in t imely manner. The system is bu i l t in a modular way, incorporat ing several existing algori thms and packages. There are three basic funct ional components to S K I C A T , serving the purposes of sky object catalog construct ion, catalog management, and high-level stat ist ical and scientific analysis.