AUTHORS · BUGS · SEE_ALSO

## NAME

irlincor -- Correct IR imager frames for non-linearity.

## USAGE

`irlincor input output`

## PARAMETERS

- input
- The list of images to be corrected for non-linearity

- output
- The list of corrected output images

- coeff= 1.0
- The first coefficient of the correction function

- coeff= 0.0
- The second coefficient of the correction function

- coeff= 0.0
- The third coefficient of the correction function

## DESCRIPTION

The IR imager frames specified by *input*
, which may be a general image
template including wild cards or an @list, are corrected for non-linearity
on a pixel by pixel basis and written to *output*
. The number of output
images must match the number input. The pixel type of the output image(s) will
match that of the input image(s), however, internally all calculations are
performed as type real. The correction is performed assuming
that the non-linearity can be represented by the following simple relationship:

ADU' = ADU * [ coeff1 + coeff2 * (ADU / 32767) + coeff3 * (ADU / 32767)**2 ]The coefficients which occur in this expression are specified by the parameters

*coeff1*,

*coeff2*and

*coeff3*. Their values are derived from periodic instrumental calibrations and are believed to be fairly constant. The default values specify a

**null**correction. You should consult

**Jay Elias**for the latest values. Note that the coefficients are expressed in terms of ADU normalised to the maximum possible value 32767, in order that their values can be input more easily.

## EXAMPLES

1. Correct input to output using the default values for the coefficients (not a very rewarding opperation!)

cl> irlincor input output

2. Correct a list of images in place using specified values for the coefficients

cl> irlincor @list @list coeff1=1.0 coeff2=0.1 coeff3=0.01

## TIME REQUIREMENTS

## AUTHORS

The IRLINCOR task was originally written by Steve Heathcote as part of the CTIO package.

## BUGS

The form of the correction equation is currently experimental; a higher order polynomial or a different functional form could be accomodated very easily if required. It may be advisable to carry out the calculations in double precision.

## SEE ALSO

onedspec.coincor, proto.imfunction