For transforms to or from sky coordinates, imagexform uses the spacecraft attitude at the midpoint of the image exposure. The transformed image will be inaccurate if the spacecraft moved significantly during the exposure.
The "events" method is appropriate if the input image represents photon events. This method mimics the action of the coordinator tool. For each event, it chooses a random position of the event within its pixel and then determines the location of that event in the transformed image. This randomness eliminated artifacts due to the misalignment of the pixels in the two images. The transformed image will have the same total number of counts as the input image, unless some of the events fell outside the transformed image.
The "intensity" method is appropriate for other types of images (e.g. counts/s). For each pixel in the transformed image, it determines the corresponding pixel in the input image. The transformed pixel is assigned a value bi-linearly
The "center" method assigns the value of an input pixel to the output pixel containing the transformed input pixel center. If multiple input pixels are mapped to the same output pixel, their values are summed.
The "area" method distributes the value of the input pixel among the output pixels that are overlapped by the quadrilateral defined by transforming the corners of the input pixel. The input value is distributed according to the fractional overlap area.
1. run imagexform prompting for options
imagexform2. run imagexform specifying parameters on the command line
imagexform infile=IMAGE_RAW.fits outfile=IMAGE_SKY.fits \ teldef=SAT_TELDEF.fits from=RAW to=SKY