Enhanced statistical stability in coherent interferometric imaging

Josselin Garnier (1) , L. Borcea , G. Papanicolaou , C. Tsogka

1 LPMA - Laboratoire de Probabilités et Modèles Aléatoires

We analyze the resolution and statistical fluctuations of images when the ambient medium is random and scattering can be modeled primarily by wavefront distortion. We compare the coherent interferometric imaging method to the widely used Kirchhoff migration and show how the latter loses statistical stability at an exponential rate with the distance of propagation. In Kirchhoff migration we form images by superposing the array data back-propagated to the image domain. In coherent interferometry, we back-propagate local cross-correlations of the array data. This is a denoising process that enhances the signal-to-noise ratio of images but also reduces the resolution. We quantify analytically the trade-off between enhanced stability and reduced resolution in coherent interferometric imaging.

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Josselin Garnier, L. Borcea, G. Papanicolaou, C. Tsogka. Enhanced statistical stability in coherent interferometric imaging. Inverse Problems, 2011, 27 (8), pp.085004. ⟨10.1088/0266-5611/27/8/085004⟩. ⟨hal-00660326⟩