Nonparametric methods for detecting change in Multitemporal SAR/PolSAR Satellite Data

• DOI: 10.48550/arxiv.2001.05764
• Publication Date: 2020-01-01
• Authors: Rodney V. Fonseca; Guilherme Ludwig; Michel H. Montoril; Aluísio Pinheiro
• Keywords: FOS: Computer and information sciences; G.3; I.4.9; G.3; I.4.9; Applications (stat.AP); 0101 mathematics; 62G07, 62M10; Statistics - Applications; 01 natural sciences

We employ nonparametric statistical procedures to analyse multitemporal SAR/PolSAR satellite images. The aim is two-fold. seek parsimony in data representation as well efficient change detection. For these, wavelets and geostatistical analyses are applied the images (Morettin et al., 2017; Krainski 2018). Following this representation, dimension of underlying generating process estimated (Fonseca Pinheiro, 2019), a set multivariate characteristics extracted. Change-points then detected via (Montoril 2019).