Binary Slice Fit Ellipticity Analysis

Abstract

Many astronomical objects show intensity sections that approximate an elliptical shape. Optimal estimates of the parameters of the ellipse best-fit of the various sections are needed in the quantitative astrophysics of galaxies, globular star clusters, and other extended objects. The methods used come from the morphological characterization of the objects. The ellipse best-fit of isophotes is used for objects, like galaxies, that approximate the linear combination of an extended continuum to a quasi-normal noise (1). Current methods for objects, like globular star clusters, that show a substantial amount of non-normal noise intrinsic to their discrete-like morphology are based on interactive techniques involving the human operator, like star counting, and visual, photographic, or isodensitometric ellipse contour fit (2,3,4,5). Of course, star counting is limited by spatial resolution and visual-supported contour fit affected by personal bias, while being in any case highly time consuming. More effective impersonal methods can be defined using numerical image processing. One simple implementation based on the iterative fit-reject technique is described in the following. It was realized for the automated analysis of globular star clusters and proved to yield results satisfactorily similar to the visual estimates when comparable, while being able to manage without relevant problems cases of low aignal to noise ratio.