Montage: Astronomical Image Mosaics, Examination, and Visualization

Montage 7.0 adds two major capabilities. The first is a set of tools for building HiPS map. HiPS (Hierarchical Progressive Surveys) is a hierarchical tiling mechanism which allows one to access, visualize and browse seamlessly image data and in particular large-scale, high-resolution surveys.

HiPS construction through Montage consists of building large-scale mosaics using pre-existing Montage modules (reprojection, background matching, and coaddition) with a HiPS-specific projection (HPX). The image hierarchy just requires repetative shrinking of higher resolution images by factors of two. Finally, the tiles to be served are simply 512x512 cutouts from these mosaics.

For high-resolution data, this process can benefit greatly from massive parallelization and this can be achieved in a number of ways. In particular, tools have been developed to streamline this on cloud platforms like AWS.

So at arcminute scale all-sky processing can adequately be done on a single desktop machine and at arcsecond scale the same tools can create a set of jobs that can be submitted to run on a cloud in a few days (or a few tens of processors if one has that in-house).

The second addition to Montage is a complete set of modern procedures for building Python binary extension Montage wheels for Linux and Mac systems (and extensible to some others). This is very much a moving target and we are building as many of these as we can and pushing them to PyPI but if someone want to extend Montage for their own use they can use the same infrastructure to build custom wheels as well.

Montage (http://montage.ipac.caltech.edu) is an Open Source toolkit, distributed with a BSD 3-clause license, for assembling Flexible Image Transport System (FITS) images into mosaics according to the user's custom specifications of coordinates, projection, spatial sampling, rotation and background matching.

The toolkit contains utilities for reprojecting and background matching images, assembling them into mosaics, visualizing the results, and discovering, analyzing and understanding image metadata from archives or the user's images.

Montage is written in ANSI-C and is portable across all common Unix-like platforms, including Linux, Solaris, Mac OSX and Cygwin on Windows. The package provides both stand-alone executables and the same functionality in library form. It has been cross-compiled to provide native Windows executables and packaged as a binary Python extension (available via "pip install MontagePy").

The distribution contains all libraries needed to build the toolkit from a single simple "make" command, including CFITSIO and the WCS library (which has been extended to support HEALPix and World-Wide Telescope TOAST projections. The toolkit is in wide use in astronomy to support research projects, and to support pipeline development, product generation and image visualization for major projects and missions; e.g. Spitzer Space Telescope, Herschel, Kepler, AKARI and others. Montage is used as an exemplar application by the computer science community in developing next-generation cyberinfrastructure, especially workflow frameworks on distributed platforms, including multiple clouds.

Montage provides multiple reprojection algorithms optimized for different needs, maximizing alternately flux conservation, range of projections, and speed.

The visualization module supports full (three-color) display of FITS images and publication quality overlays of catalogs (scaled symbols), image metadata, and coordinate grids. It fits in equally well in pipelines or as the basis for interactive image exploration and there is Python support for the latter (It has also been used in web/Javascript applications).

We are in the process of adding automated regression testing using Jenkins. At the moment, this only includes a couple of dummy tests on a Jenkins server that we maintain specifically for the Montage project).

Montage was funded from 2002 to 2005 by the National Aeronautics and Space Administration's Earth Science Technology Office, Computation Technologies Project, under Cooperative Agreement Number NCC5-626 between NASA and the California Institute of Technology. The Montage distribution includes an adaptation of the MOPEX algorithm developed at the Spitzer Science Center. Montage has also been funded by the National Science Foundation under Award Number NSF ACI-1440620.