Manpage of imblob
Section: User Commands (1)
Return to Main Contents
imblob - Fit 2-D gaussian in an image or cube
IMBLOB is a Miriad task which fits a 2D gaussian to a user
specified region. Only intensities larger than a specified
cutoff are fit.
The model fit is :
G(x,y) = OFFSET + AMP * exp(-(x-X0)**2/(2*A**2) +
where A>B, and the x,y coordinate system is rotated by
PA degrees w.r.t. the NAXIS2 axis (astronomical convention)
towards the positive NAXIS1 axis.
A and B are the SIGMA, such that
FWHM = sqrt(8*ln2) * SIGMA = 2.355 * SIGMA
The input image dataset name.
Region over which the gaussian is to be fitted. Full region
specifications are supported.
Cutoff below which pixels are excluded from the fit. All points
included are weighted equally. Use a large enough negative number
to include all points.
If need be, this parameter can be used to supply the fitting routine
with initial conditions other than the ones it derived on its own.
The order of the parameters is: FWHM1, FWHM2 (arcsec), PA (degrees),
AMP, OFFSET, XCEN, YCEN (absolute pixel coordinates: 1..naxisN).
Default uses the derived values.
If you want to fix the center, enter two numbers XCEN, YCEN here.
They must be in absolute pixel coordinates. Default: not used.
Note: using this keyword overrides the XCEN, YCEN used in ini=
put will be for 2D gaussians:
Channel nr, Velocity, Xcen, Ycen(ARCSEC-REL), Xcen, Ycen (PIXEL-abs),
AMP, FWHM_major, FWHM_minor, PA (degrees),
- PERSON RESPONSIBLE
This document was created by
using the manual pages.
Time: 18:35:38 GMT, July 05, 2011