A Method of Polarimetric Synthetic Aperture Radar Image Classification Based on Sparse Representation

Abstract

IntroductionFeature extraction and PolSAR image interpretation are of much theoretical and application significance, and recently, the extensive attention in PolSAR remote sensing has gradually arisen. PolSAR related researches have been conducted for many years, and a variety of methods have been proposed. Many of the proposed techniques are based on physical scattering mechanisms obtained from different polarimetric decomposition methods. Polarimetric target decomposition theorem expresses the average mechanism as the sum of independent elements in order to associate a physical mechanism with each resolution cell, which allows the identification and separation of scattering mechanisms in polarization signature for purposes of classification and recognition. Many descriptive parameters can be extracted by coherent and incoherent decomposition methods based on polarimetric characteristics. Freeman and Durden suggested a three-component decomposition method of polarimetric data, which introduced a combination of surface, double-bounce, and volume scatterings.Based on the idea of scattering model decomposition, a variety of incoherent decomposition models were proposed, such as four-component scattering model and multiplecomponent scattering model (MCSM) [1][2][3][4][5][6][7][8].The PolSAR image classification is actually a high-dimensional nonlinear mapping problem. The extensive analysis of the physical mechanism is difficult, and for complex scenes, the underlying physical mechanism of each pixel is hard to obtain. Thus, the high-level image processing and learning machine are useful for the classification of the complex scenes. Sparse representation of signals based on over-complete dictionary is a kind of new signal representation theory, which substitutes over-complete redundant function system for traditional orthogonal basis functions and provides great flexibility for adaptive sparse extension of signals [9][10][11][12][13]. It aims to approximate a target signal using a linear combination of elementary signals from a large candidate set, which is called as "dictionary", with each element called as "atom". Sparse representation has therefore increasingly become recognized as providing extremely high performance for diverse applications. The sparse representation-based target classification is a kind of new theory in SAR images processing, and has been proved to be rather efficacious [14].

Compact Power Divider Integrated with Coupler and Microstrip Cavity Filter for X-band Surveillance Radar System

“…Surveillance Radar and Polarimetric Synthetic Aperture Radar [4] is a technology that is need for an archipelagic nation like Indonesia. Other than that, antenna suitable for coastal surveillance radar is looking at a beam of the antenna, as in previous research circularly symmetric shaped beams [5].…”