Ship Discrimination Using Polarimetric SAR Data and Coherent Time-Frequency Analysis

Abstract

This paper presents a new approach for the discrimination of ship responses using polarimetric SAR (PolSAR) data. The PolSAR multidimensional information is analyzed using a linear Time-Frequency (TF) decomposition approach, which permits to describe the polarimetric behavior of a ship and its background area for different azimuthal angles of observation and frequencies of illumination. This paper proposes to discriminate ships from their background by using characteristics of their polarimetric TF responses, which may be associated with the intrinsic nature of the observed natural or artificial scattering structures. A statistical descriptor related to polarimetric coherence of the signal in the TF domain is proposed for detecting ships in different complex backgrounds, including SAR azimuth ambiguities, artifacts, and small natural islands, which may induce numerous false alarms. Choices of the TF analysis direction, i.e., along separate azimuth or range axis, or simultaneously in both directions, are investigated and evaluated. TF decomposition modes including range direction perform better in terms of discriminating ships from range focusing artifacts. In comparison with original full-resolution polarimetric indicators, the proposed TF polarimetric coherence descriptor is shown to qualitatively enhance the ship/background contrast and improve discrimination capabilities. Using polarimetric RADARSAT-2 data acquired over complex scenes, experimental results demonstrate the efficiency of this approach in terms of ship location retrieval and response characterization.

Polarimetric time-frequency analysis of vessels in Spotlight SAR images

“…Thus, for a maritime scenario it is likely that ships, or at least some of the scatterers from the ship, exhibit a high coherency indicator. Conversely the sea will exhibit low values, as already shown successfully on full polarimetric data in [9].…”

“…Some points of the target are still classified as non-stationary; this effect could be due to relevant azimuth orientation (i.e. anisotropic target behavior) but it may be caused also by processing artifacts [9]. In Figure 3 an optical image of Gibraltar is shown (upper panel), along with the correlation coefficient between HH and VV channel (ρ HHV V , lower left panel) and the coherency indicator (lower right panel), both relative to a portion of the represented area.…”

Power Transmission Tower Series Extraction in PolSAR Image Based on Time-Frequency Analysis and A-Contrario Theory

“…TF analysis of the sub-band spectrum independently in range and azimuth directions not only can be used for the detection of anisotropic targets (including scatterers with complex geometry, artificial targets and linear arranged strong scatterers) and targets whose scattering characteristics are sensitive to frequency (such as cylinder targets, periodic structure or coupled scatterers), but can also be used to find non-stationary regions in distance-azimuth spectrums [19,20]. In this paper, the TF decomposition is based on two-dimensional (2D) short-term Fourier Transform [21]. …”

Built-up Area Extraction from PolSAR Imagery with Model-Based Decomposition and Polarimetric Coherence

“…The sub-aperture or time-frequency decomposition, which uses a two-dimensional Fourier transform, or Gabor transform to decompose a full resolution image into several sub-aperture images, has been widely used on SAR image application [36][37][38][39][40]. A 2D signal s (l) can be transformed into several different components located around particular spectral coordinates l p ; ω p using a convolution with an analyzing function g () as follows: where l p and ω p denote spatial and frequency locations of the pth component, respectively.…”