Abstract

"Restoration" is the term used in the image processing field for the operation of "deconvolution" (inversion of the linear integral equation of the first kind). In ground-based optical astronomy the convolved function is the "seeing", the approximately Gaussian point-source response which we see when we image an unresolved star through the turbulent atmosphere. Obviously observers would like to "restore" the image details which are attenuated by the convolution. The terms "seeing" (in optical astronomy), "beam" (in radio astronomy), "instrumental profile" (in spectroscopy), "point spread function", and "kernel" are used interchangeably to refer to the convolving kernel of the integral equation of imaging. The term "reconstruction" is sometimes used as a synonym for restoration, but more often it is used in situations where the data are integrals along projections in various directions and the analyst wants to reconstruct the object which probably produced the data. There are strong analogies between reconstruction and restoration problems and workers in either field will generally find some profit in watching the activities in the other field.

© 1983 Optical Society of America