WCSCOPY - Copies WCS information from one NDF to another

Description:

This application copies the WCS  component from one NDF  to another, optionally modifying it to take account of a linear mapping between the pixel co-ordinates  in the two NDFs. It can be used, for instance, to rectify the loss of WCS information produced by older applications which do not propagate the WCS component.

Usage:

wcscopy ndf like [tr] [confirm]

Parameters:

CONFIRM = _LOGICAL (Read)

If TRUE, the user is asked for confirmation before replacing any existing WCS component within the input NDF. No confirmation is required if there is no WCS component in the input NDF. [TRUE]

LIKE = NDF (Read)

The reference NDF data structure from which WCS information is to be copied.

NDF = NDF (Read and Write)

The input NDF data structure in which the WCS information is to be stored. Any existing WCS component is over-written (see parameter CONFIRM).

OK = _LOGICAL (Read)

This parameter is used to get a confirmation that an existing WCS component within the input NDF can be over-written.

TR( ) = _DOUBLE (Read)

The values of this parameter are the coefficients of a linear transformation from pixel co-ordinates in the reference NDF given for Parameter LIKE, to pixel co-ordinates in the input NDF given for Parameter NDF. For instance, if a feature has pixel co-ordinates (X,Y,Z,...) in the reference NDF, and pixel co-ordinates ($U$ ,$V$ ,$W$ ,...) in the input NDF, then the following transformations would be used, depending on how many axes each NDF has:

• one-dimensional:
  $$U = TR(1) + TR(2)*X$$

• two-dimensional:
  $$U = TR(1) + TR(2)*X + TR(3)*Y$$
  $$V = TR(4) + TR(5)*X + TR(6)*Y$$

• three-dimensional:
  $$U = TR(1) + TR(2)*X + TR(3)*Y + TR(4)*Z$$
  $$V = TR(5) + TR(6)*X + TR(7)*Y + TR(8)*Z$$
  $$W = TR(9) + TR(10)*X + TR(11)*Y + TR(12)*Z$$

If a null value (!) is given it is assumed that the pixel co-ordinates of a given feature are identical in the two NDFs. [!]

Examples:

wcscopy m51_sim m51

This copies the WCS component from the NDF called m51 to the NDF called m51_sim, which may hold the results of a numerical simulation for instance. It is assumed that the two NDFs are aligned (i.e. the pixel co-ordinates of any feature are the same in both NDFs).

wcscopy m51_sqorst m51 [125,0.5,0.0,125,0.0,0.5]

This example assumes that an application similar to SQORST has previously been used to change the size of a two-dimensional NDF called m51, producing a new NDF called m51_sqorst. It is assumed that this SQORST-like application does not propagate WCS and also resets the pixel origin to [1, 1]. In fact, this is what SQORST actually did, prior to KAPPA version 1.0. This example shows how WCSCOPY can be used to rectify this by copying the WCS component from the original NDF m51 to the squashed NDF m51_sqorst, modifying it in the process to take account of both the squashing and the resetting of the pixel origin produced by SQORST. To do this, you need to work out the transformation in pixel co-ordinates produced by SQORST, and specify this when running WCSCOPY using the TR parameter. Let's assume the first axis of NDF m51 has pixel-index bounds of I1:I2 (these values can be found using NDFTRACE). If the first axis in the squashed NDF m51_sqorst spansM pixels (where M is the value assigned to SQORST Parameter XDIM), then it will have pixel-index bounds of 1:M. Note, the lower bound is 1 since the pixel origin has been reset by SQORST. The squashing factor for the first axis is then:

$$FX = M/(I2 - I1 + 1)$$

and the shift in the pixel origin is:

$$SX = FX*( 1 - I1 )$$

Likewise, if the bounds of the second axis in m51 are J1:J2, and SQORST Parameter YDIM is set to N, then the squashing factor for the second axis is:

$$FY = N/(J2 - J1 + 1)$$

and the shift in the pixel origin is:

$$SY = FY*( 1 - J1 )$$

You would then use the following values for Parameter TR when running WCSCOPY:

$$TR = [SX, FX, 0.0, SY, 0.0, FY]$$

Note, the zero terms indicate that the axes are independent (i.e. there is no rotation of the image). The numerical values in the example are for an image with pixel-index bounds of 52:251 on both axes which was squashed by SQORST to produce an image with 100 pixels on each axis.

Notes:

• An error is reported if the transformation supplied using Parameter TR is singular.

• The pixel with pixel index I spans a range of pixel co-ordinate from $(I - 1.0)$ to I.

• The pixel indices of the bottom-left pixel in an NDF is called the pixel origin of the NDF, and can take any value. The pixel origin can be examined using application NDFTRACE and set using application SETORIGIN. WCSCOPY takes account of the pixel origins in the two NDFs when modifying the WCS component. Thus, if a null value is given for Parameter TR, the supplied WCS component may still be modified if the two NDFs have different pixel origins.

Related Applications :

KAPPA: NDFTRACE, WCSADD, WCSATTRIB, WCSFRAME, WCSREMOVE.