OBJMASKing methods


The first OBJMASK ("Primary") is intended to eliminate the brighest stars with a large block size (-5) and large hsigma (2)
With these large blocks we do not get a good mask around edges of galaxies where the light is on a gradient, but for bright stars, this works well.

This "Primary" mask has the follow parameters

images =	Core.fits	List of images or MEF files
objmasks =	CoreP2.pl	List of output object masks
omtype =	numbers	Object mask type
skys =		List of input/output sky maps
sigmas =		List of input/output sigma maps
masks =	RMCore.pl	List of input bad pixel masks
extname =		Extension names
logfile =	STDOUT	List of log files
blkstep =	1	Line step for sky sampling
blksize =	-5	Sky block size (+=pixels, -=blocks)
convolv =		Convolution kernel
hsigma =	2	Sigma threshold above sky
lsigma =	4	Sigma threshold below sky
hdetect =	yes	Detect objects above sky?
ldetect =	yes	Detect objects below sky
neighbo =	4	Neighbor type
minipix =	2	Minimum number of pixels in detected objects
ngrow =	1	Number of grow rings
agrow =	1	Area grow factor
mode =	a

(In subsequent mentions of OBJMASKing, only the blksize and hsigma values will be cited as the other parameters remained constant.)

Example: /Core/CoreP2.pl     blksize -5, hsigma 2


Now that the main stars and large portions of galaxies have been masked, we employ a more rigorous mask to get at the rest of the smaller stars and remaining galaxies.
In this mask, we use blksize -20, hsigma 1. Smaller blocks allow for better masking across galaxy fringes as they take more localised views of the sky around them.

Example: /Core/CoreS7.pl     blksize -20, hsigma 1


back to main Virgo page