Hue-saturation-intensity split-spectrum processing of Seasat radar imagery

Characteristics of terrain features in orbital radar imagery are investigated on the basis of radar physics and the human visual system. Signal scattering, the dominant factor in radar imagery, is caused by look angle and relief displacement. Power spectra characteristics are explored, noting that surface scattering dominates the low frequencies and slope significantly affects the high frequency tone variations. The maximum human visual response has been identified for achromatic scenes as 7 cycles per degree, whereas in radar imagery the low frequency half-power point occurs near 2.5 cycles/deg. Image feature size at a scale of 1:500,000 has been calculated for the maximum human response at 1.25 km. A hue-saturation-intensity transformation is presented for displaying two or three data sets on the same image and is demonstrated for Seasat SAR imagery. The transformation is shown to restore low spatial frequencies to the SAR imagery in order to identify terrain features which would otherwise not be visually discernible.