A new classifier for Polarimetric SAR (PolSAR) images is proposed and assessed in this paper. Its input consists of segments, and each one is assigned the class which minimizes a stochastic distance. Assuming the complex Wishart model, several stochastic distances are obtained from the h-φ family of divergences, and they are employed to derive hypothesis test statistics that are also used in the classification process. This article also presents, as a novelty, analytic expressions for the test statistics based on the following stochastic distances between complex Wishart models: Kullback-Leibler, Bhattacharyya, Hellinger, Rényi, and Chi-Square; also, the test statistic based on the Bhattacharyya distance between multivariate Gaussian distributions is presented. The classifier performance is evaluated using simulated and real PolSAR data. The simulated data are based on the complex Wishart model, aiming at the analysis of the proposal well controlled data. The real data refer to the complex L-band image, acquired during the 1994 SIR-C mission. The results of the proposed classifier are compared with those obtained by a Wishart per-pixel/contextual classifier, and we show the better performance of the region-based classification. The influence of the statistical modeling is assessed by comparing the results using the Bhattacharyya distance between multivariate Gaussian distributions for amplitude data. The results with simulated data indicate that the proposed classification method has a very good performance when the data follow the Wishart model. The proposed classifier also performs better than the per-pixel/contextual classifier and the Bhattacharyya Gaussian distance using SIR-C PolSAR data.

No abstract

Given the different nature of optical and radar data, it is reasonable the idea that each type of data can contribute in complementary ways for different applications. This paper aims at analyzing the potential joint usage of optical and Synthetic Aperture Radar (SAR) data for land use and land cover classification in a region located in the Brazilian Amazon. To achieve this objective, we evaluated regionbased classifications using separated and fused optical and SAR data. Data were images from the Landsat 5/TM sensor and amplitude multipolarized images from the ALOS/PALSAR sensor. The images were classified using a region-based classifier based on the Bhattacharyya distance between Gaussian distributions. The TM data alone is better for classify land cover classes with occurrence of trees or shrubs, while SAR data contribute to improve the classification results in low vegetated areas.

Region-based classification of PolSAR data can be effectively performed by seeking for the assignment that minimizes a distance between prototypes and segments. Silva et al. (2013) used stochastic distances between complex multivariate Wishart models which, differently from other measures, are computationally tractable. In this work we assess the robustness of such approach with respect to errors in the training stage, and propose an extension that alleviates such problems. We introduce robustness in the process by incorporating a combination of radial basis kernel functions and stochastic distances with Support Vector Machines (SVM). We consider several stochastic distances between Wishart: Bhatacharyya, Kullback-Leibler, Chi-Square, Rényi, and Hellinger. We perform two case studies with PolSAR images, both simulated and from actual sensors, and different classification scenarios to compare the performance of Minimum Distance and SVM classification frameworks. With this, we model the situation of imperfect training samples. We show that SVM with the proposed kernel functions achieves better performance with respect to Minimum Distance, at the expense of more computational resources and the need of parameter tuning. Code and data are provided for reproducibility.

This work presents a region based classifier for Polarimetric SAR (PolSAR) images. The classifier uses the stochastic distances derived from the complex Wishart Model, obtained from the h-φ family of divergences. Adittionaly, a hypothesis test derived from the stochastic distance is also employed in the classification process. The region based classifier, using the Bhattacharyya distance, was applied to a polarimetric SIR-C image from an agricultural area in northeastern Brazil. The region based classification result significantly overperformed the a pixel based/contextual PolSAR classification based on the Maximum Likelihood/Iterated Conditional Modes. Such evidence lead us to conclude that the region based stochastic distance and hypothesis test classifier offers a good potential at identifying the land cover classes on a PolSAR image.