Glossary for Geospatial Science

Fourier transform:  A type of two-way frequency transform for identifying and removing unwanted spatial frequency components in an image.  The Fourier process is normally used to identify and remove systematic noise “spikes,” such as regular lines that may have been introduced by a faulty image collection device.  The process works as a two-way operation.  First the forward transform creates a pair of intermediate raster objects that reveal abnormal data variability.  After the values in those raster objects have been edited to remove the data spikes, the inverse transformation creates a raster object with the same image as the original, but with lines or other noise removed.  This Fourier Transform process has somewhat the same smoothing effect as the low-frequency TNTmips image filtering process.