A wavelet regularization method for diffusion radar-target imaging and speckle noise reduction (Q1318617): Difference between revisions

The theory and applications of wavelets have become a very active field of research in the last few years, because they are standing at the intersection of the frontiers of mathematics, algorithmic and non- algorithmic computing, signal processing, and remote sensing. Radar imagery can be performed by recording and displaying a mapping of the reflectivity of target surfaces or terrains [\textit{V. C. Chen}, Radar ambiguity function, time-varying matched filter, and optimum wavelet correlator. Optical Eng. 33, 2212-2217 (1994)]. The paper under review deals with radar imagery under a diffuse-target statistical model for the scattering off a rough target surface. The radar image is the scattering function which describes the second-order statistics of target reflectivity in planar (delay, Doppler) coordinates. Based on a wavelet representation for the logarithm of the scattering function instead of the scattering function itself, the problem of restoring a radar image is regarded as a problem of inference on the wavelet coefficients of the image corrupted by additive non-Gaussian noise. In this approach significant components of the scattering function are captured at different levels of resolution. The author studies a hypothesis testing technique for solving the problem at a singificance level prescribed by the desired noise/reduction trade- off. The statistical model for the image restoration problem applies also for synthetic aperture radar (SAR) imagery and optical computing [\textit{E. N. Leith}, Synthetic aperture radar. In \textit{D. Casasent} (ed.), Optical data processing, 89-117 (1978); \textit{J. P. Fitch}, Synthetic aperture radar. (1988); \textit{J. Detlefsen}, Radartechnik. (1989); \textit{D. Blacknell}, The resampling of synthetic aperture radar images. In \textit{S. R. Brooks} (ed.), Mathematics in remote sensing, 173-188 (1989; Zbl 0705.62097)]. Thus, similar to spline techniques, wavelet methods also have an impact on the stochastic inverse problems of statistical signal processing such as ISAR.