It summarizes related techniques selleck chem inhibitor on SAR statistical modeling before 1997. After 1997, papers on SAR statistical modeling have appeared in renowned journals almost every year. The most attractive achievement among them is the statistical modeling on extremely heterogeneous region of SAR images proposed by Frery [24], who works in Brazil and has introduced the original idea that for the purpose of statistical modeling, SAR images can be divided into homogeneous regions, heterogeneous regions and extremely heterogeneous regions, according to their contents. Furthermore, statistical modeling of SAR images is taken as one of the main contents in more than 20 doctoral dissertations found in UMI and in the research reports from the Belgian Royal Military Academy.

While numerous statistical distributions Inhibitors,Modulators,Libraries have been proposed to model Inhibitors,Modulators,Libraries SAR image data, we are unaware of any surveys on this particular topic. It is necessary to categorize and evaluate these models and relevant issues. The main contribution of this survey is the classification and evaluation of the statistical models of SAR images existed currently. The vital and latest contributions have also been covered in this paper. The survey is organized as follows: Section 2 illustrates the classification and the research contents of statistical modeling. In Sections 3 and 4, current statistical models are discussed in detail. The relationship of them and their limitations in applications are pointed out in Section 5. Major conclusions and developing trends of statistical modeling Inhibitors,Modulators,Libraries are also presented by Section 6.

We conclude the survey in the final section.2.?Model Classification and Research ContentsAccording to the modeling process, the statistical models of SAR images can be divided into Inhibitors,Modulators,Libraries two categories [2,25�C28]: parametric models and nonparametric models. When dealing with a parametric model, several known probability distributions of SAR imagery are given at first. Usually, the parameters of these distributions are unknown and have to be estimated according to the real image data. Finally, by using some certain metrics, the optimal distribution is chosen as the statistical model of the AV-951 image. While handling a nonparametric model, no distributions have to be assumed, and the optimal distribution is obtained in a way of data-driven of image data.

The merit selleck catalog of the nonparametric models is that they make the process of statistical modeling more flexible and can fit the real data more accurate.Since nonparametric modeling involves complex computation as well as numerous data, it is usually time-consuming and cannot satisfy the requirements of various applications [25]. Consequently, parametric modeling is intensively studied. The process of parametric modeling can be described in brief as to choose an appropriate one from several given statistical distributions for the image to be modeled. The process is shown in Figure 1.