xnregistar: Mosaic the images using sub-pixel replication and masking

Package: xdimsum

Usage

xnregistar inlist rmasks output expmap sections

Parameters

inlist: The input shiftlist normally produced by the xmshifts, xfshifts, xrshifts, or xdshifts tasks. Inlist contains the name of the input sky subtracted image, the x offset, the y offset, and the exposure time in columns 1 through 4.

rmasks: The list of input rejection masks normally produced by the xmskcombine task. The rjection mask is assumed to consist of 1's and 0's where 1's define the good values. Rrmasks may define a list masks which has the same length as inlist, or an image keyword containing the name of the rejection mask. The default keyword name written by xmskcombine is "REJMASK". If no rejection mask is defined a virtual cosmic ray mask consisting entirely of 1's is defined.

output: The name of the output combined image.

expmap: The name of the output exposure map image.

sections: The optional output sections or corners file suitable for input to the maskdereg task. If defined sections contains the name of the input images in sinlist if it is defined or inlist, the x and y coordinates of the lower left corner of the input image in the combined output image, and the x and y coordinates of the upper right corner of the input image in the output image in columns 1 through 5 respectively.

sinlist = "": The list of image names that will be written to the sections file. If sinlist is undefined then the input image names in inlist are used.

nprev_omask = 0: The number of previous object masks to combine to create an objects mask. By default no object masks are defined. If nprev_omask is 1 then the object mask of the previous image is used to mask the current image, if nprev_omask = 2 then the 2 previous object masks are used, etc.

blkrep = yes: Use block replication to magnify the image if mag > 1 rather than bilinear interpolation ?

mag = 1: The default magnification factor. If mag = 1 the scale of the output image is the same as the scale of the input images. If mag > 1 then the input images are block replicated by a factor of mag before being combined to create the output images.

fractional = no: If fractional is yes then the input images are shifted by the fractional part of the total offset in inlist before being combined. If fractional is no then the fractional part of the pixel shift is lost.

pixin = yes: If pixin is yes the input offsets called a and b are assumed to be defined in the reference image pixel coordinate system x and y and ab_sense, xscale, yscale, and a2x_angle default to yes, -1.0, -1.0, and 0.0 respectively.

ab_sense = yes: Is the rotation of the a to b axis counter-clockwise ? The x and b axes are the axes along which the input offsets are measured. Ab_sense defaults to yes if pixin is yes.

xscale = 1.0, yscale = 1.0: The number of pixel coordinates x and y per input coordinates a and b. For example if the input offsets are measured in arcseconds and the pixel scale is 0.25 arcseconds per pixel then xscale and yscale should be set to 4.0. Xscale and yscale default to -1.0 and -1.0 if pixin is yes.

a2x_angle = 0: The angle in degrees of the a offset axis measured counter-clockwise to the image x axis. A2x_angle defaults to 0.0 is pixin is yes.

ncoavg = 1: The number of co-averages per input image. The total exposure time for an image is ncoavg * exptime * secpexp where exptime is read from inlist.

secpexp = 1.0: The number of seconds per unit exposure time.

y2n_angle = 0.0: The angle in degrees from the image y axis to the north axis measured from north through east. Y2n_angle can be used to orient the output image to within 45 degrees of N up and E left if set correctly.

rotation = yes: Is the rotation of the input image north axis to the east axis counter-clockwise ?

Description

XNREGISTAR uses the image names, positional offsets, and exposure time data in inlist to create a combined output image output and associated exposure map image expmap. Bad data in each input image is excluded from the output image and exposure map image using the rejection masks rmasks. Rejection masks consist entirely of 1's and 0's with 1's defining the good pixels. Rejection masks are normally created by the XMSKCOMBINE task which combines information in the global bad pixel mask, the cosmic ray masks (normally written by the XZAP or XNZAP tasks), the input image holes mask (normally written by the XSLM task), and the object masks (normally written by the MKMASK and MASKDEREG tasks), to create a single rejection mask for each input image. If the sections file sections is defined an output file describing the location of each input image in the output image is also written. The sections file is suitable for input to the MASKDEREG task.

The input image names in column 1 of inlist are the names of the sky subtracted images normally written by the XSLM task. If sinlist is "" these names are written to the sections file sections. Otherwise the names in sinlist are used. The number of images in sinlist must be 0 or the same as the number of input images.

The offsets in columns 2 and 3 of inlist are assumed to be defined in the ab coordinate system. If pixin = yes then the ab coordinate system is assumed to be the same as the xy coordinate system of the reference image used to compute the offsets. If pixin = no then the ab_sense, xscale, yscale, and a2x_angle are used to determine the transformation from the ab to the xy coordinate system. If mag is > 1 then the input images are block replicated by a factor of mag before being combined into the output image and output exposure map image.

The integer pixel offsets for each input image are computed as shown below. These offsets are passed directly to the IMCOMBINE task which does the actual image combining. The quantities a and b are the coordinates as read directly from inlist.

tmag = mag
if (pixin == yes) {
    ab_sense = yes
    txscale = -1.0
    tyscale = -1.0
    ta2x = 0.0
    sign = 1
} else if (ab_sense == yes) {
    txscale = xscale
    tyscale = yscale
    ta2x = a2x_angle
    sign = 1
} else {
    txscale = xscale
    tyscale = yscale
    ta2x = a2x_angle
    sign = -1
}

x = tmag * (txscale * a * cos (ta2x) +
    tyscale * b * sign * sin (ta2x))
y = tmag * (tyscale * b * sign * cos (ta2x) -
    txscale * a * sin (ta2x))
if (x == 0.0)
    ix = 0
else
    ix = int (x + 0.5 * (x / abs (x)))
if (y == 0.0)
    iy = 0
else
    iy = int (y + 0.5 * (y / abs (y)))

If fractional = yes then the input images are shifted by the fractional part of the pixel offsets using the IMSHIFTS task before being combined with the IMCOMBINE task. If fractional = no the fractional part of the shift is ignored. The fractional shifts are defined as follows.

xfrac = newx - ix
yfrac = newy - iy

The input images are multiplied by a factor equal to ncoavg / (mag * mag) before being summed into the combined output image output. The summed output image is divided by the exposure map image to compute the final output image. Each input image contributes a factor equal to texp * ncoavg * secpexp to the exposure map image, where texp is the exposure time read from inlist.

The bad data masks bpmask, crmasks, holes, and omasks are used to exclude bad pixels, known cosmic ray pixels, and undefined pixels from the combined output image and output exposure map image.

Finally if y2n_angle and rotation are set correctly the output image and the output exposure map image will be oriented to within 45 degrees of north pointing up and east pointing left.

Examples

1. Compute the total offsets for the series of 25 demo sky subtracted images whose total offsets were computed with the xdshifts task. Combine the images with xnregister using the bad pixel mask demo.pl. The output image and exposure map image will be written to demo.mosaic and exp_demo.mosaic respectively.

cl> type simlist
demo01.sub.im
demo02.sub.imh
demo03.sub.imh
...
...
demo24.sub.imh
demo25.sub.imh

cl> xdshifts @simlist "" offsets cradius=5.0

cl> xmskcombine @simlist badpix.pl "" "" "" ".rjm"

cl> xnregistar offsets REJMASK  demo.mosaic demo.mosaic.exp ""

xnregistar: Mosaic the images using sub-pixel replication and masking

Usage

Parameters

Description

Examples

Time requirements

Bugs

See also