Images

Images

In LSST, 2-dimensional images are represented using a set of templated classes, all of which are ultimately derived from ImageBase:

• Image : An image of arithmetic type (e.g. short; int; float)
• Mask : An image representing a set of bitplanes
• DecoratedImage : An Image that has associated metadata (e.g. from a FITS header)
• MaskedImage : A combination of an Image, a Mask, and an Image giving the per-pixel variance
• Exposure : A MaskedImage with additional WCS information

You can use ds9 to display Images and their Masks.

How to Access Pixels

Image Pixels may be accessed via iterators or locators; the former are simpler if you want single-pixel access, while the latter provide you with access to a pixel's friends and neighbours. The following tutorials are available:

• Image iterators
• Image locators
• MaskedImage iterators
• MaskedImage locators

See also Pixel Access Reference Manual. LSST image access is modelled after the boost::gil image interface; see http://www.boost.org/doc/libs/1_48_0/libs/gil/doc/html/giltutorial.html and http://www.boost.org/doc/libs/1_48_0/libs/gil/doc/html/gildesignguide.html

• Image I/O

There are also some algorithms analogous to the STL's <algorithm>; see How to use algorithms to manipulate Images.

Masks

• How Mask Planes are handled in afw