Representing backscatter fluctuations with a PDF convolution equation and its application to study backscatter variability in side-scan sonar images

Image histograms are useful for revealing the relative proportions of geological components in a scene, but these proportions are commonly obscured by the variability caused by signal fading (image speckle). This variability can potentially be reduced by deconvolution because a scene's PDF is related to the Rayleigh log-PDF and PDF of backscatter strengths via a convolution integral. A simple deconvolution algorithm is demonstrated by separating a submarine lava flow from its adjacent sediments in a side-scan sonar image histogram.