Back to Volume
Paper: Image Stacking Tools for Modern Surveys
Volume: 394, Astronomical Data Analysis Software and Systems (ADASS) XVII
Page: 493
Authors: Neilsen, E.H., Jr.
Abstract: Modern imaging observation programs often collect many exposures for each area of sky observed. Although simple methods for generating a single, high quality exposure from collections of overlapping images are well understood and tools that implement them are commonplace, many modern exposure sets, such as those from the Sloan Digital Sky Survey (SDSS) and those expected from the Dark Energy Survey (DES), have complicating properties that these tools do not address optimally. These exposures may have different point spread functions, so direct coaddition or image stacking will not result in an image with either an optimal PSF or noise. They may have significantly different distortion in their mapping of pixel to celestial coordinates. If the generated image is to be used for object detection, exposures through different filters should be combined into a single image, so that objects below the detection threshold in any given filter may still be detected. I present progress on a set of tools designed to combine multiple exposures in a manner optimal for object detection, integrating distortion removal and registration, coaddition, and noise suppression into a single step. I also introduce a method for using cross correlation between exposures to estimate the statistical properties of undetected sources for use in combining images.
Back to Volume