Previous: Target acquisition tools
Up: LEXT manual
Next: Data processing
Previous Page: Target acquisition tools
Next Page: Data processing

Mapping focal plane to detector coordinates

There are two stages to the mapping between the mask focal plane coordiante system and the detector coordinates. The first is the linear mapping of pixels at the detector to mm at the mask, and the second is to account for the small optical distortion introduced into the detector coordinates by the LDSS-2 optics.

Although generally small, field distortion is can be used for use with PREBOX and with mask manufacture from direct images. It is not used in the extraction routines which have algorithms for following distortions empirically.

The important routines are:

• LFIT fits an optimised 5 parameter linear transform to a pair of list slots (ie. stretch, translation and rotation). The main function of this routine is to define the first order mapping from mask coordinates (in mm) to detector coordinates (in pixels). The coefficients are displayed.

  If the /STORE qualifier is present then it is assumed that the fit defines the mapping of the mask coordinates onto the detector and resets the values of the relevant internal parameters accordingly. This allows the subsequent use of the TRANSFORM command.

• SFIT fits a bicubic spline to a pair of list slots. It thus maps the distortion between the two sets of points as a function of position. It uses a KNOTS list to define the fiducial points for which the spline parameters are calculated and stored.

  This routine can be used to define the distortion of a matrix mask image (which has been analysed by, say, MAP), compared to the expected image (see below). The fit is currently stored in an external transformation file. It can be subsequently used by the routine DISTORT.

• CREATE/GRID produces a list, according to user defined criteria of grid points such as would define a matrix mask.
• TRANSFORM transforms the list coordinates between the mask () and detector (pixels) systems and vice-versa. It produces a general, linear transformation including rotation, translation etc.

  The transformation is defined by the internal variables ROTATION, XPIX, YPIX, FRED, XCENTRE and YCENTRE.

• EVALUATE converts a transformation, as generated by SFIT, file into a 2D representation. Specifically the and shifts are stored in separate images which are given the XSHFT and YSHFT.

  If these are available then they will be used by DISTORT in preference to the raw spline coefficients, since it saves on CPU time.

• DISTORT applies a spline transformation, such as generated in SFIT, to a list or image. By default a fractional pixel interpolation scheme is employed, although this can be over-ridden by other qualifiers.