Detection and Measurement of Poorly Sampled Point Sources Imaged With 2-D Array

Abstract

Imaging systems in which the pixels are large compared to the point spread function produce undersampled data for which traditional 2-D Gaussian PSF fitting will not work well. Such systems include wide-field imaging applications (CCD mosaics) and space-based telescopes. The current astronomical literature provides few recipes to use when dealing with undersampled data. We present a method of providing optimum signal-to-noise data matching for poorly sampled point sources which makes use of profile fitting but only within small, variable-size pixel masks. A wide-field imaging Schmidt telescope project, Lowell Observatory Near-Earth Object Search (LONEOS), is discussed as an example. Our pixel mask technique is applied to model images from the LONEOS camera, and we show that we can determine point-source centroids and brightnesses with good precision, even for faint objects.