Sky’s No Limit for Computer: Software system catalogs, classifies stars and galaxies

Telescopes can photograph hundreds of millions of objects in the sky, but the task of trying to classify and catalog the objects by hand would make any astronomer faint at heart.

And so the computer once again is called upon to perform one of astronomy’s more laborious tasks.

Scientists at Caltech in Pasadena and the Jet Propulsion Laboratory in La Canada Flintridge have developed a computer software system to help them analyze nearly 3,000 photographic piates of the northern sky, captured by telescopes at Mount Palomar.

Their software has one important difference: It “learns” how to classify different types of galaxies, much like a human being, only better, said Caltech astronomy graduate student Nick Weir.

Not only is the system – dubbed SKICAT, for Sky Image Cataloging and Analysis Tool – free from hman failings like fatigue, it also can classify stars and galaxies too faint for humans to recognize.

“SKICAT represents a new generation of intelligent, trainable tools for dealing with the huge volumes of scientific image data that today’s instruments collect,” said JPL scientist Usama Fayyad, who helped develop the software.

As images from Palomar’s northern sky survey roll in, a facility in Boston will process them, then send them to Caltech for SKICAT to work its cataloging and classifying magic.

SKICAT, funded by NASA, will be publically available, Weir said, and plans are in the works to share the software with other astronomical institutions.

Eventually, as technology gallops along, images from a telescope will be immediately processed and cataloged at the observatory. But that’s a few years down the road, Weir said.

“Meanwhile, SKICAT will serve a big gap,” he said.

By: Elizabeth Wilson

Source: Pasadena Star News

PDF

Leave a Reply