Getting Started

Documentation

Glish

Learn More

Programming

Contact Us

Version 1.9 Build 1367

News	FAQ
Search	Home

Introduction

The Imaging Model as recommended for AIPS++ in the report of the Green Bank Meeting (Cornwell and Shone, 1992) is somewhat briefly and incompletely described in that document. This document proposes a more formal definition of the Imaging Model and of the various terms used in the GB report and subsequently in AIPS++ design documents.

The purpose of the Imaging Model was to act as an abstraction of how Telescopes convert Sky Brightness into Data. All Telescopes which conform to this abstraction would then be suitable for processing with AIPS++ with relatively little coding. All that would be needed in many cases would be a specialized form of the Imaging Model class. Actually the Calibration model would be needed as well but I totally ignore all issues connected with calibration in this document. To the extent that this abstraction is correct, we can therefore avoid duplication of effort in writing deconvolution tasks.

On reading the literature of image deconvolution, especially in radio astronomy, it is notable that many seemingly different approachs have been tried over the years. This is discouraging for a project like AIPS++ in which we hope to be able to apply many different types of deconvolution algorithm to data from many different telescopes. Despite this variety, I demonstrate below that one scheme can account for many different types of telescope. This scheme is based upon one particular abstraction of a telescope: namely, the way it converts images into data. The scheme is supported by two things: first, a reading of image processing literature outside of radio astronomy, and second, experience with SDE in which some of these ideas have been tried out. The implications of this scheme should be most apparent in the complicated imaging methods investigated by Holdaway and Bhatnagar (1992). I believe that some simplification of their suggested procedures may result.