__________________________________________________________________________________________________________ Proc. of CEOS Working Group on Calibration / Validation SAR Workshop, London, United Kingdom, 24 – 26 September 2002 (ESA SP-526, March 2003) Natural Resources Canada 588 Booth Street, Ottawa Ontario, Canada K1A OY7 Abstract — Cameron’s coherent target decomposition and classification are discussed in the context of It is shown that Cameron’s classification leads to a coarse scattering segmentation because of the large class dispersion that corresponds to a SAR system with about ±5 dB channel The of Cameron’s A new method, named the symmetric scattering characterization method (SSCM), is introduced to exploit the information provided by the largest target symmetric scattering com- ponent, under coherent conditions. The SSCM, which ex- pressed the symmetric scattering in term of the Poincar´ e sphere angles, permits a better characterization of target symmetric scattering and the generation of coherent scatter- ing segmentation of much higher resolution, in comparison with Cameron’s coarse segmentation. I. Introduction An interesting classification method, which is based on coherent target decomposition (CTD), was introduced in [1]. Cameron’s classification extracts, in each illuminated resolution cell, the largest symmetric scattering component from the total radar return, and assigns it to one of six sym- metric elemental scatterer classes: the trihedral, diplane, dipole, cylinder, narrow diplane, and quarter wave device [1]. This method has been widely used for the characteriza- tion and identification of point targets such as ships [7], [8] and small planes. In Section II, Cameron’s CTD and classi- fication are discussed in the context of SAR, and the limita- tions of Cameron’s classification are examined. Cameron’s CTD is then reconsidered in Section III, to develop a new method, named the SSCM method, that exploits better the information provided by Cameron’s CTD in the con- text of coherent scattering. The new method is validated for ocean scattering characterization and ship identifica- tion using Convair-580 polarimetric SAR data collected off shore Cape Race, Newfoundland, Canada. * under contract with TGIS Consultant II. Analysis of Cameron’s Coherent Target Decomposition and classification A. Cameron’s CTD Inspired by the work of Huynen [4], Cameron associates importance to a class of targets termed symmetric. A sym- metric target as defined in [4] is a target having an axis of symmetry in the plane orthogonal to the radar line of sight direction (LOS). Symmetric targets have a scattering ma- trix which can be diagonalized by a rigid rotation about the LOS in a basis of linear eigen polarizations. Huynen intro- duced a technique for CTD in terms of target parameters that are orientation invariant, and particularly in terms of the polarization configurations that maximizes the received power in [4]. The ellipticity angle τ m of the maximum po- larization determines the symmetric-nonsymmetric nature of the scattering; τ m = 0 for symmetric scattering and τ m = 0 for asymmetrical scattering. Cameron developed an algorithm that maximizes the symmetrical component of coherent scattering [1]. Under target and SAR system reciprocity assumption, the scat- tering matrix is expressed in terms of the Pauli matrices, as: [S]= α[S a ]+ β[S b ]+ γ [S c ]. Scattering is symmetric if there exists an angle of rotation ψ a that cancels the projec- tion of [S] on the non symmetric Pauli direction  S c , where  S c is the vectorial form of the Pauli matrix [S c ]. This leads to the following expression for the symmetric part,  S sym , as a function of the angle θ = −2ψ a [1]:  S sym = α  S a + ǫ · [cos θ ·  S b + sin θ ·  S c ] (1) The symmetric component  S sym of the total scattering  S (the vectorial form of [S]), reaches its maximum for the angle θ that satisfies the following relationship, for β = γ [1]: tan(2θ)= βγ ∗ + β ∗ γ imbalance. ments limits the utility of the classification. application method within known SAR radiometric calibration require- better (CTD) SAR, and the limitations of Cameron’s classification are ex- amined. R. Touzi and F. Charbonneau* Canada Centre for Remote Sensing THE SSCM: AN ADAPTATION OF CAMERON'S TARGET DECOMPOSITION TO ACTUAL CALIBRATION SAR REQUIREMENTS