GoogleSky: the Paper

Sky in Google Earth: The Next Frontier in Astronomical Data Discovery and Visualization

This paper has just been released on the Astrophysics Preprint Service. It's the first step to teach users, and in particular astronomers, how to add data to GoogleSky directly from their own images. It will likely have a large impact on the community, since from now anyone will be able to take their own data (in FITS format) and generate a proper KML with a single command.

In fact, the paper contains a link to open source software some Google developers released that allows for FITS to KML conversion. More precisely, it allows for a FITS image with a proper WCS (World Coordinate System) to be converted into KML in two way:
  • A simple translation of the input FITS image into a single PNG that is stitched to the sphere that is GoogleSky
  • A translation of the input FITS image into a series of "regionated" PNG file, properly warped onto the GoogleSky sphere. This approach allows to serve large collections of data with only small chunks at a time.
This software is called WCS2KML and it is written in python and C++

At STScI, we have lots of experience in python and C++ and we are planning on taking this software and turning it into a pipeline tool for astronomers.
Unfortunately, not everyone can add to the open source repository as the Google developers want to maintain some control over the software. This will practically mean that STScI will branch from this current version.

In particular:

- Python code depends on Numeric. Numeric is dead. It was first replaced by numarray and now by numpy. Even the Numeric people don't support it anymore. Code would need to be upgraded to numpy before broad distribution.

- Should really use pyfits instead of their pyfits lite application.

- WCS2KML only supports png format. This seems limiting at the moment.

No comments:

Post a Comment