A Statistical Approach to Monitor Earth's Surface in Remote Sensing Applications

Abstract

In Earth's surface monitoring, the most significant signature of the target is the scattering mechanism, i.e., α-angle, whose evaluation requires special attention and solution. In these investigations, the α-angle possesses statistical features depending on the type of the scattering. There are several methods, such as target decomposition, eigenvector analysis, and the maximum likelihood estimator, to recognize the target in natural environments. In this article, the combination of target decomposition and maximum likelihood estimator is addressed as a new algorithm to investigate radar targets. It will be demonstrated that several probability density functions, such as Rayleigh, normal, gamma, and binomial, can be used to estimate the α-angle. To validate analytical results, polarimetric synthetic aperture radar (PolSAR) data, provided by the European Space Agency, are investigated. The consequences justify the potential of the proposed algorithm.