MUSE Pipeline Reference Manual  2.1.1
Functions
2D Source Detection

Functions

cpl_table * muse_find_stars (const cpl_image *aImage, double aHmin, double aFwhm, const double *aRoundLimits, const double *aSharpLimits)
 Find positive brightness perturbations (i.e stars) in an image. More...
 

Detailed Description

This group provides a simple 2d image source detection based of the DAOFIND algorithm (Stetson P.B., 1987, PASP 99, 191).

This implementation of the DAOFIND algorithm is largely based on the C implementation of the IDL find routine which was developed by A. Pecontal to be used in the MUSE Instrument Control Software.

Function Documentation

cpl_table* muse_find_stars ( const cpl_image *  aImage,
double  aHmin,
double  aFwhm,
const double *  aRoundLimits,
const double *  aSharpLimits 
)

Find positive brightness perturbations (i.e stars) in an image.

Parameters
aImage2D image to be searched for stars.
aHminMinimum value above background for threshold detection
aFwhmFWHM (in pixels) to be used in the convolution filter
aSharpLimits2 element vector giving low and high cutoff for the sharpness statistic.
aRoundLimits2 element vector giving low and high cutoff for the roundness statistic.
Returns
On success, a table containing the pixel coordinates and the measured properties of the detected sources is returned, or NULL in case of an error.

Find positive brightness perturbations (i.e stars) in an image and return the centroid position, (uncalibrated) flux and the shape parameters (roundness and sharpness).

For each detected star the measured properties are stored in one row of the results table. The measured coordinates, flux, sharpness and roundness are written to the table columns X, Y, Flux, Sharpness, and Roundness, respectively.

The size of the input image aImage must be even along its x-axis. The minimum value above the background aHmin should, in general, be set to 3 or 4 sigma above background RMS. The input FWHM aFwhm must be larger than 0.5 pixels.

If NULL is passed for aSharpLimits and/or aRoundLimits, the built-in defaults are used.

The default ranges for aSharpLimits and/or aRoundLimits are [0.2,1.0] and [-1.0,1.0] respectively. The limits of the sharpness statistic should only be changed if the stars have a significantly smaller or larger concentration than a Gaussian. Similarly, the limits of the roundness statistic should only be changed if the stars are significantly elongated.

Adapted from IDL implementation of the DAOPHOT FIND routine which does not allow for bad pixels and uses a slightly different centroid algorithm. Uses marginal Gaussian fits to find centroids.

Note
  • The sharpness statistic compares the central pixel to the mean of the surrounding pixels. If this difference is greater than the originally estimated height of the Gaussian or less than 0.2 the height of the Gaussian (for the default values of aSharpLimits) then the star will be rejected.
  • use marginal Gaussian distributions to compute centroid As discussed in more detail in the comments to the code, the centroid computation here is the same as in IRAF DAOFIND but differs slightly from the current DAOPHOT.
Exceptions
set CPL_ERROR_NULL_INPUT, return NULLa parameter is NULL
set CPL_ERROR_ILLEGAL_INPUT, return NULLa parameter is not a valid input
set CPL_ERROR_INCOMPATIBLE_INPUT, return NULLconvolution of the input image with a Gaussian failed
set CPL_ERROR_DATA_NOT_FOUND, return NULLno candidate sources where found

Definition at line 455 of file muse_findstars.c.

References muse_convolve_image(), muse_cplarray_erase_invalid(), muse_cplmatrix_extract_selected(), muse_cplmatrix_multiply_create(), and muse_cplmatrix_where().