This paper deals with the electromagnetic scattering of plane wave by a randomly rough surface. Scattering mechanism is extracted in backscattered alignment or mono-static radar, made up of a non periodic mesh array of finite number of rectangular sloped surfaces. By the Kirchhoff approximation, the total scattering components are calculated for several different numbers of slopes, relative dielectric constants, correlation length, and height deviation of surface and incident angle. Then eigenvector target decomposition method (EV-TDT) is applied for calculating covariance matrix eigenvectors and then estimate of entropy and scattering mechanism and then target classification. As results, for SAR data sets which supported from European space agency (ESA) data provider, polarimetric signatures of randomly rough dielectric surface such as entropy and scattering mechanism, extracted and their results are compared to theoretical results.