The CLEAN Algorithm

The CLEAN algorithm assumes that the radio sky can be represented by a small number of point sources in an otherwise empty field of view. It uses a simple iterative procedure to find the positions and strengths of these sources. The final deconvolved `CLEAN' image is the sum of:

1. These point-source `CLEAN components' re-convolved ('restored') with a `CLEAN beam' (usually Gaussian) to de-emphasize the higher spatial frequencies which are often spuriously extrapolated and,
2. (optionally, but strongly recommended) an image representing residual differences between the point-source model and the data.

This algorithm proceeds as follows:

1. Find the strength and position of the peak (i.e., of the greatest absolute intensity) in the dirty image.
2. Subtract from the dirty image, at the position of the peak, the dirty beam B multiplied by the peak strength and a damping factor $\gamma$ (usually termed the loop gain).
3. Go to (1) unless all remaining peaks are below some user-specified level. The search for peaks may be constrained to specified areas of the image, called `CLEAN' windows.
4. Convolve the accumulated point source model with an idealized `CLEAN' beam (usually an elliptical Gaussian fitted to the central lobe of the dirty beam).
5. Add the residuals of the dirty image to the `CLEAN' image.

(Ref: Image Reconstruction, Prof. Dale E. Gary , NJIT
(http://web.njit.edu/$\sim$dgary/728/Lecture7.html))