Processing of Synthetic Aperture Radar Images

Although it is relatively recent, radar interferometry technique is already widely used by the entire radar remote sensing community. The principle was described by Graham in 1974 and validated on SEASAT and SIR-B data. However, it is only since the launching of the first European radar satellite ERS-1 in 1991 that it has rapidly developed. Since then, data acquisition by tandem ERS 1 & 2 (two images obtained by two sensors during a one-day interval) and, more recently, by the SRTM mission (two antennae placed on the shuttle) have enabled the implementation of this technique on a planetary scale.
Interferometry uses phase information contained in radar images as a very sensitive means of distance measurement, or more exactly of distance variations, because it makes sense only in a differential manner. Finite analysis of phase difference between two radar images makes it possible to reconstitute land forms or to measure land movements that may have taken place between several data collections with millimetric precision.
Currently, many laboratories use interferometry (or simply phase coherence, which is its sub-product) for their applications; these applications themselves are numerous in geophysics, topography, geology, glaciology, hydrology, volcanology, forestry, gas or oil exploitation, or the management of mineral resources in general. [1]
Each pixel of a complex radar image involves two sets of information: radiometry (wave amplitude) and phase.
Radiometry represents ground reflectivity to microwave: this value constitutes stricto sensu the image. Phase is the sum of two terms: