一种自适应的混合Freeman/Eigenvalue极化分解模型

An adaptive hybrid Freeman/Eigenvalue polarimetric decomposition model

全极化SAR数据的极化分解在土地利用分类、目标检测与识别以及地表参数反演等领域得到了广泛应用。目前,主要有基于特征值分解和基于模型分解2类极化分解方法。混合Freeman/Eigenvalue极化分解结合了两者的优势,避免了基于模型的极化分解中负功率问题并且能够利用已知的散射机制解释分解后的散射分量。为了进一步拓展该分解在不同地表类型中的应用,通过引入参数Neumann一般化体散射模型,提出了一种自适应的极化分解模型。利用德国Black Forest地区的L波段AirSAR(airborne synthetic aperture Radar)全极化数据进行实验,并与现有的Yamaguchi三分量模型和自适应非负分解(adaptive nonnegative eigenvalue decomposition,ANNED)对比分析,以验证模型的有效性。研究表明,自适应的混合Freeman/Eigenvalue极化分解模型保证了分解能量的非负性及完全分解,适应于不同类型的地表,能有效地区分不同地类。

Polarimetric decomposition of fully polarimetric SAR data has been extensively used in land use classification, target detection and identification, and land surface parameter retrieval. At present, two main categories of polarimetric decomposition approaches can be identified, i.e., model-based decomposition and eigenvalue-based decomposition. By combining the advantages of the above two decomposition methods, the hybrid Freeman/Eigenvalue method can deal with the negative power problems, and the decomposed components can be interpreted in terms of known scattering mechanisms. In order to extend the applicability of the hybrid Freeman/Eigenvalue to different types of land cover, the authors propose a novel adaptive polarimetric decomposition method in this paper by coupling the hybrid Freeman/Eigenvalue decomposition and an adaptive volume scattering model proposed by Neumann et al. The performance and advantages of the proposed method were demonstrated and evaluated with AirSAR L-band data over Black Forest in Germany. Comparative studies were also carried out with previous Yamaguchi three-component decomposition and adaptive nonnegative eigenvalue decomposition （ANNED）. The results show that the proposed method can effectively avoid negative power problems and is applicable to different types of land cover. Moreover, different types of land cover can be well discriminated by the proposed technique.