A new polarimetric classification approach evaluated for agricultural crops

Statistical properties of the polarimetric backscatter behavior for a single homogeneous area with constant radar reflectivity are described by the Wishart distribution or its marginal distributions. These distributions do not necessarily well describe the statistics for a collection of homogeneous areas of the same class because of variation in, for example, biophysical parameters. Using Kolmogorov-Smirnov (KS) tests of fit, it is shown that, for example, the Beta distribution is a better descriptor for the polarimetric correlation, and the log-normal distribution for the backscatter level. An evaluation is given for a number of agricultural crop classes, grasslands, and fruit tree plantations at the Flevoland test site, using an AirSAR (C-, L- and P-band polarimetric) image from July 3, 1991. A new reversible transform of the covariance matrix into backscatter intensities will be introduced in order to describe the full polarimetric target properties in a mathematically alternative way, allowing for the development of simple, versatile, and robust classifiers. Moreover, it allows for polarimetric image segmentation using conventional approaches. The effect of azimuthally asymmetric backscatter behavior on the classification results is discussed. Several models are proposed, and results are compared with results from the literature for the same test site. It can be concluded that the introduced classifiers perform very well, with levels of accuracy for this test site, with 14 cover types, of 90.4% for C-band, 88.7% for L-band, and 96.3% for the combination of C- and L-band.