Using position lists

In CCDPACK the positions of identified/detected image features are stored in ordinary text files which are referred to as `position lists'. The format of these lists is flexible. Usually position lists have three columns:

Identifier    X-position    Y-position

these may be separated by commas or blanks. The identifier is an integer value which is used to identify positions which are related (i.e. are of the same object) in different lists.

If more than three columns exist then only the values in the first three are used, though some of the programs propagate extra columns from the input to the output lists. If only two columns exist it is assumed that they are:

X-position    Y-position

such lists may be produced by KAPPA applications. In this case applications which rely on a knowledge of the identifiers assume they are monotonically increasing from one.

Whole and in-line comments are allowed in position lists using the character `#'. These are not propagated from input to output lists. Note that the final line of the file should be terminated by a newline character (certain text editors do not always enforce this).

The X and Y values in a position list always refer to the pixel coordinates of an image. When FINDOFF or REGISTER read them they automatically transform them into the Current image coordinate system where appropriate. For this reason there is not normally any need to transform a position lists the corresponding image has been resampled; however if necessary it can be done using the TRANLIST routine. This is discussed, along with the methods that previous versions of CCDPACK used for storing coordinate transformations, in appendix .

Usually position lists are `associated' with images. What this means is that when a position list is created a record of its name is kept in the extension of the image. It is then usual to refer to the image instead of the position list when the position list is to be accessed. Applications which create new position lists associate the new position lists with the appropriate image. Using this method avoids any confusion about the relationship of position lists and images, which is vital when determining the registration of many images at one go. It also allows the use of the wildcarding properties of image names to access position lists.

The association of position lists can be disabled completely by setting the NDFNAMES global parameter (CCDSETUP) to false. In this case, position lists must be specified explicitly as a comma separated list of names or gotten from a text file using indirection, this is exactly the same as for naming images (see ``processing lists of data'') except that wildcards are not allowed. Output list names may be formed from the modification of these names when NDFNAMES is false, otherwise the input image names are always used.

The name of the position list associated with an image is stored in its .MORE.CCDPACK extension under the item CURRENT_LIST, and can be examined using HDSTRACE (SUN/102) or modified using CCDEDIT.

If you wish to view or edit positions in a list interactively, you can use the IDICURS program, which displays a graphical interface in which the image is displayed with points from a position list plotted over it. Points can be added and removed with the mouse. If you want to display points in a list non-interactively or to print them out, use PLOTLIST.