Previous Article Next Article Table of Contents
An Algorithm for Locating the Solar Limb (1Jun92) (from NSO, NOAO Newsletter No. 30, 1 June 1992) We have developed an algorithm for measuring the location of the solar limb in full disk images that is insensitive to atmospheric seeing conditions. The algorithm also provides an estimate of the form of the point spread function for each observed image. After further testing, this algorithm will be available to the community. The location of the solar limb is important for solar oscillation measurements. Any inaccuracies in the determination of the image geometry will translate into an asymmetry in the observed ridges in the solar oscillation power spectrum. Such `analysis induced' asymmetries can obviously lead to systematic errors in any estimate of the properties of the features in the power spectrum. The information on the point spread function is necessary for the correction of variations in the observed p-mode amplitudes due to changes in atmospheric conditions (a major drawback in trying to make accurate measurements of the oscillation amplitudes). The method employed essentially consists of three steps: (1) Generate the Laplacian map of an image and use it to determine a high quality estimate of the coordinates of the center of the image and the values of the semi-major and semi-minor axes. (2) Use the property that the zero crossing points of the Fourier transform of the observed intensity profile (after appropriate edge extension to alleviate problems associated with truncation of the image signal) are related solely to the limb darkening profile. This then enables a determination of the error in the initial estimates for the semi-major and semi-minor axes and also of the form of the limb darkening function by least-squares methods. (3) Knowledge of the form of the limb darkening function makes it possible to estimate the form of the point spread function for the observation by deconvolution methods. Cliff Toner, Stuart Jefferies
Previous Article Next Article Table of Contents