基于标准Huynen二分法的极化SAR完全模型分解强化改进

doi:10.7523/j.ucas.2024.046

摘要/Abstract

摘要：

基于模型的极化分解旨在将复杂目标极化散射在表面、二面和体散射等标准模型上加性展开，是实现极化合成孔径雷达精细解译的关键技术。然而模型分解存在负功率和信息利用不充分等问题。完全模型分解（CMD）及其改进虽解决了这些问题，却进一步引发零功率退化。其表现为，CMD并不总能从体散射移除后的剩余矩阵中分解出一个表面和二面分量，零散射功率不可避免。本文对CMD进行强化改进，发现CMD对表面和二面分量的提取本质上是一个偏好于体散射的标准Huynen二分法。通过将该分解自适应升级为标准Huynen二分法中偏好于表面和二面散射的二分法，强化分解总能无退化地分解出表面和二面散射分量，从而有效解决零功率问题。在实测数据上的定性和定量对比验证了所提方法的有效性。

关键词: 极化合成孔径雷达, 极化分解, 模型分解, 散射建模, 目标二分法

Abstract:

As an important technique for the fine interpretation of the polarimetric synthetic aperture radar targets， the model-based polarimetric decomposition is dedicated to additively expanding the complex target scatterings over the canonical models of surface scattering， double-bounce scattering， and volume scattering. However， model-based decompositions suffer from negative powers and insufficient utilization of polarimetric information. The complete model-based decomposition （CMD） and its improvement solve these problems， but cause the so-called zero-power degradation. This is because CMD can not always extract a surface component and a double-bounce component from the remaining matrix after the removal of the volume scattering， and zero scattering power is then inevitable. CMD is reinforced in this paper. We find that the extraction of the surface and double-bounce components in CMD is essentially a sub-dichotomy of the canonical Huynen dichotomy that prefers volume scattering. By adaptively upgrading it with the other two sub-dichotomies of the canonical Huynen dichotomy that prefer surface scattering and dihedral scattering， respectively， the reinforced CMD can always decompose the surface and double-bounce components without degradation，and the problem of zero power is thus effectively solved. Qualitative and quantitative comparisons on measured polarimetric SAR datasets are presented to demonstrate the good performance of the proposed method.

Key words: polarimetric synthetic aperture radar, polarimetric decomposition, model-based decomposition, scattering modeling, target dichotomy

中图分类号:

TN957.52

李佳橦, 李东, 张云华, 王勋, 陆禾. 基于标准Huynen二分法的极化SAR完全模型分解强化改进[J]. 中国科学院大学学报, 2026, 43(3): 404-413.

Jiatong LI, Dong LI, Yunhua ZHANG, Xun WANG, He LU. Reinforcement of the complete model-based decomposition for polarimetric SAR based on canonical Huynen dichotomy[J]. Journal of University of Chinese Academy of Sciences, 2026, 43(3): 404-413.

图/表 11

图1 RCMD总体流程图

Fig.1 Flow chart of RCMD

图2 Oberpfaffenhofen地区ESAR机载全极化数据的地物类型参考图及Pauli RGB伪彩色图

Fig.2 Terrain feature and Pauli RGB pseudo-color image of the airborne full-polarimetric data over Oberpfaffenhofen acquired by ESAR

图3 San Francisco地区GF-3机载全极化数据的地物类型参考图及Pauli RGB伪彩色图

Fig.3 Terrain feature and Pauli RGB pseudo-color image of the airborne full-polarimetric data over San Francisco acquired by GF-3

图4 Oberpfaffenhofen地区不同分解方法的结果

Fig.4 Results obtained by different decomposition methods over Oberpfaffenhofen

表1 Oberpfaffenhofen地区归一化散射功率平均值 (%)

Table 1 Mean normalized scattering power over Oberpfaffenhofen

方法	$P s$	$P d$	$P v$
CMD-ED	45.97	26.79	27.24
CMD-MF	45.43	27.33	27.24
ICMD	45.57	27.19	27.24
RCMD	44.40	28.37	27.23

表1 Oberpfaffenhofen地区归一化散射功率平均值 (%)

Table 1 Mean normalized scattering power over Oberpfaffenhofen

方法	$P s$	$P d$	$P v$
CMD-ED	45.97	26.79	27.24
CMD-MF	45.43	27.33	27.24
ICMD	45.57	27.19	27.24
RCMD	44.40	28.37	27.23

表2 Oberpfaffenhofen地区正功率像素比例 (%)

Table 2 Percentage of positive power pixels over Oberpfaffenhofen

方法	$P s$	$P d$	$P v$
CMD-ED	98.88	96.42	100
CMD-MF	93.38	77.67	100
ICMD	93.96	77.46	100
RCMD	100	100	100

表2 Oberpfaffenhofen地区正功率像素比例 (%)

Table 2 Percentage of positive power pixels over Oberpfaffenhofen

方法	$P s$	$P d$	$P v$
CMD-ED	98.88	96.42	100
CMD-MF	93.38	77.67	100
ICMD	93.96	77.46	100
RCMD	100	100	100

表3 Oberpfaffenhofen地区各区域不同分解方法提取的各分量功率的占比 (%)

Table 3 Percentage of each component power extracted by different decomposition methods in different subregions over Oberpfaffenhofen

区域	方法	$P s$	$P d$	$P v$
A	CMD-ED	16.99	29.97	53.04
	CMD-MF	18.23	28.73	53.04
	ICMD	18.34	28.62	53.04
	RCMD	17.80	29.19	53.01
B	CMD-ED	27.26	56.38	16.36
	CMD-MF	28.65	54.99	16.36
	ICMD	28.61	55.03	16.36
	RCMD	27.10	56.55	16.35
C	CMD-ED	76.25	14.46	9.29
	CMD-MF	76.62	14.09	9.29
	ICMD	76.67	14.04	9.29
	RCMD	76.72	13.99	9.29

表3 Oberpfaffenhofen地区各区域不同分解方法提取的各分量功率的占比 (%)

Table 3 Percentage of each component power extracted by different decomposition methods in different subregions over Oberpfaffenhofen

区域	方法	$P s$	$P d$	$P v$
A	CMD-ED	16.99	29.97	53.04
	CMD-MF	18.23	28.73	53.04
	ICMD	18.34	28.62	53.04
	RCMD	17.80	29.19	53.01
B	CMD-ED	27.26	56.38	16.36
	CMD-MF	28.65	54.99	16.36
	ICMD	28.61	55.03	16.36
	RCMD	27.10	56.55	16.35
C	CMD-ED	76.25	14.46	9.29
	CMD-MF	76.62	14.09	9.29
	ICMD	76.67	14.04	9.29
	RCMD	76.72	13.99	9.29

图5 San Francisco地区不同分解方法的结果

Fig.5 Results obtained by different decomposition methods over San Francisco

表4 San Francisco地区归一化散射功率平均值 (%)

Table 4 Mean normalized scattering powers over San Francisco

方法	$P s$	$P d$	$P v$
CMD-ED	70.58	14.49	14.93
CMD-MF	71.37	13.70	14.93
ICMD	71.56	13.51	14.93
RCMD	69.99	16.23	14.78

表4 San Francisco地区归一化散射功率平均值 (%)

Table 4 Mean normalized scattering powers over San Francisco

方法	$P s$	$P d$	$P v$
CMD-ED	70.58	14.49	14.93
CMD-MF	71.37	13.70	14.93
ICMD	71.56	13.51	14.93
RCMD	69.99	16.23	14.78

表5 San Francisco地区正功率像素百分比 (%)

Table 5 Percentage of positive power pixels over San Francisco

方法	$P s$	$P d$	$P v$
CMD-ED	98.72	94.62	100
CMD-MF	97.70	35.58	100
ICMD	98.74	33.97	100
RCMD	100	100	100

表5 San Francisco地区正功率像素百分比 (%)

Table 5 Percentage of positive power pixels over San Francisco

方法	$P s$	$P d$	$P v$
CMD-ED	98.72	94.62	100
CMD-MF	97.70	35.58	100
ICMD	98.74	33.97	100
RCMD	100	100	100

表6 San Francisco地区各区域不同分解方法提取的各分量功率的占比 (%)

Table 6 Percentage of each component power extracted by different decomposition methods in different subregions over San Francisco

区域	方法	$P s$	$P d$	$P v$
A	CMD-ED	98.40	0.77	0.83
	CMD-MF	98.78	0.39	0.83
	ICMD	98.82	0.35	0.83
	RCMD	99.16	0.01	0.83
B	CMD-ED	37.13	28.41	34.46
	CMD-MF	40.80	24.74	34.46
	ICMD	40.25	25.29	34.46
	RCMD	37.28	28.56	34.06
C	CMD-ED	40.36	15.32	44.32
	CMD-MF	39.15	16.53	44.32
	ICMD	41.95	13.73	44.32
	RCMD	38.19	17.96	43.85

表6 San Francisco地区各区域不同分解方法提取的各分量功率的占比 (%)

Table 6 Percentage of each component power extracted by different decomposition methods in different subregions over San Francisco

区域	方法	$P s$	$P d$	$P v$
A	CMD-ED	98.40	0.77	0.83
	CMD-MF	98.78	0.39	0.83
	ICMD	98.82	0.35	0.83
	RCMD	99.16	0.01	0.83
B	CMD-ED	37.13	28.41	34.46
	CMD-MF	40.80	24.74	34.46
	ICMD	40.25	25.29	34.46
	RCMD	37.28	28.56	34.06
C	CMD-ED	40.36	15.32	44.32
	CMD-MF	39.15	16.53	44.32
	ICMD	41.95	13.73	44.32
	RCMD	38.19	17.96	43.85

参考文献 36

[1]	Cloude S R， Pottier E. A review of target decomposition theorems in radar polarimetry［J］. IEEE Transactions on Geoscience and Remote Sensing， 1996， 34（2）： 498-518. DOI： 10.1109/36.485127 .
[2]	Lee J S， Pottier E. Polarimetric radar imaging： from basics to applications［M］. Boca Raton， Florida， USA： CRC Press， 2017.
[3]	Cloude S. Polarisation applications in remote sensing［M］. Oxford： Oxford University Press， 2010.
[4]	王春乐，禹卫东. Huynen类型目标分解方法的比较与分析［J］. 中国科学院研究生院学报， 2011， 28 （3）： 402-409. DOI： 10.7523/j.issn.2095-6134.2011.3.019 .
[5]	Cloude S R， Pottier E. An entropy based classification scheme for land applications of polarimetric SAR［J］. IEEE Transactions on Geoscience and Remote Sensing， 1997， 35（1）： 68-78. DOI： 10.1109/36.551935 .
[6]	Yajima Y， Yamaguchi Y， Sato R， et al. POLSAR image analysis of wetlands using a modified four-component scattering power decomposition［J］. IEEE Transactions on Geoscience and Remote Sensing， 2008， 46（6）： 1667-1673. DOI： 10.1109/TGRS.2008.916326 .
[7]	Li D， Lu H， Zhang Y H. Solid angle geometry-based modeling of volume scattering with application in the adaptive decomposition of GF-3 data of sea ice in Antarctica［J］. Remote Sensing， 2023， 15（12）： 3208. DOI： 10.3390/rs15123208 .
[8]	Touzi R， Boerner W M， Lee J S， et al. A review of polarimetry in the context of synthetic aperture radar： concepts and information extraction［J］. Canadian Journal of Remote Sensing， 2004， 30（3）： 380-407. DOI： 10.5589/m04-013 .
[9]	Bargiel D. A new method for crop classification combining time series of radar images and crop phenology information［J］. Remote Sensing of Environment， 2017， 198： 369-383. DOI： 10.1016/j.rse.2017.06.022 .
[10]	Xu F， Jin Y Q. Deorientation theory of polarimetric scattering targets and application to terrain surface classification［J］. IEEE Transactions on Geoscience and Remote Sensing， 2005， 43（10）： 2351-2364. DOI： 10.1109/TGRS.2005.855064 .
[11]	Singh G， Yamaguchi Y， Boerner W M， et al. Monitoring of the March 11， 2011， off-tohoku 9.0 earthquake with super-tsunami disaster by implementing fully polarimetric high-resolution POLSAR techniques［J］. Proceedings of the IEEE， 2013， 101（3）： 831-846. DOI： 10.1109/JPROC.2012.2230311 .
[12]	Yamaguchi Y. Disaster monitoring by fully polarimetric SAR data acquired with ALOS-PALSAR［J］. Proceedings of the IEEE， 2012， 100（10）： 2851-2860. DOI： 10.1109/JPROC.2012.2195469 .
[13]	Dong H W， Zhang L M， Zou B. Exploring vision transformers for polarimetric SAR image classification［J］. IEEE Transactions on Geoscience and Remote Sensing， 2022， 60： 1-15. DOI： 10.1109/TGRS.2021.3137383 .
[14]	刘杉，张风丽，韦诗莹，等. 基于极化分解组合的SAR图像视觉优化和建筑物损毁评估［J］. 中国科学院大学学报， 2020， 37（6）： 750-759. DOI： 10.7523/j.issn.2095-6134.2020.06.005 .
[15]	王懿泽，孙吉利，闫成杰，等. 基于超像素与LightGBM的极化SAR图像地物分类［J］. 中国科学院大学学报， 2023， 40（5）： 658-669. DOI： 10.7523/j.ucas.2022.023 .
[16]	Freeman A， Durden S L. A three-component scattering model for polarimetric SAR data［J］. IEEE Transactions on Geoscience and Remote Sensing， 1998， 36（3）： 963-973. DOI： 10.1109/36.673687 .
[17]	Yamaguchi Y， Moriyama T， Ishido M， et al. Four-component scattering model for polarimetric SAR image decomposition［J］. IEEE Transactions on Geoscience and Remote Sensing， 2005， 43（8）： 1699-1706. DOI： 10.1109/TGRS.2005.852084 .
[18]	Cui Y， Yamaguchi Y， Yang J， et al. On complete model-based decomposition of polarimetric SAR coherency matrix data［J］. IEEE Transactions on Geoscience and Remote Sensing， 2014， 52（4）： 1991-2001. DOI： 10.1109/TGRS.2013.2257603 .
[19]	Chen S W， Sato M. General polarimetric model-based decomposition for coherency matrix［C］//2012 IEEE International Geoscience and Remote Sensing Symposium. Munich， Germany. IEEE， 2012： 99-102. DOI： 10.1109/IGARSS.2012.6351627 .
[20]	An W T， Cui Y， Yang J. Three-component model-based decomposition for polarimetric SAR data［J］. IEEE Transactions on Geoscience and Remote Sensing， 2010， 48（6）： 2732-2739. DOI： 10.1109/TGRS.2010.2041242 .
[21]	Sato A， Yamaguchi Y， Singh G， et al. Four-component scattering power decomposition with extended volume scattering model［J］. IEEE Geoscience and Remote Sensing Letters， 2012， 9（2）： 166-170. DOI： 10.1109/LGRS.2011.2162935 .
[22]	van Zyl J J， Arii M， Kim Y. Model-based decomposition of polarimetric SAR covariance matrices constrained for nonnegative eigenvalues［J］. IEEE Transactions on Geoscience and Remote Sensing， 2011， 49（9）： 3452-3459. DOI： 10.1109/TGRS.2011.2128325 .
[23]	刘高峰，李明，王亚军，等. 一种新的基于非反射对称非负特征值分解的Freeman分解［J］. 电子与信息学报， 2013， 35（2）： 368-375. DOI： 10.3724/SP.J.1146.2012.00897 .
[24]	Wang C L， Yu W D， Wang R， et al. Comparison of nonnegative eigenvalue decompositions with and without reflection symmetry assumptions［J］. IEEE Transactions on Geoscience and Remote Sensing， 2014， 52（4）： 2278-2287. DOI： 10.1109/TGRS.2013.2259177 .
[25]	An W T， Xie C H. An improvement on the complete model-based decomposition of polarimetric SAR data［J］. IEEE Geoscience and Remote Sensing Letters， 2014， 11（11）： 1926-1930. DOI： 10.1109/LGRS.2014.2313955 .
[26]	Li D， Zhang Y H. Unified Huynen phenomenological decomposition of radar targets and its classification applications［J］. IEEE Transactions on Geoscience and Remote Sensing， 2016， 54（2）： 723-743. DOI： 10.1109/TGRS.2015.2464113 .
[27]	Touzi R. Target scattering decomposition in terms of roll-invariant target parameters［J］. IEEE Transactions on Geoscience and Remote Sensing， 2007， 45（1）： 73-84. DOI： 10.1109/TGRS.2006.886176 .
[28]	Singh G， Yamaguchi Y， Park S-E. General four-component scattering power decomposition with unitary transformation of coherency matrix［J］. IEEE Transactions on Geoscience and Remote Sensing， 2013， 51（5）： 3014-3022. DOI： 10.1109/TGRS.2012.2212446 .
[29]	Yamaguchi Y， Sato A， Boerner W M， et al. Four-component scattering power decomposition with rotation of coherency matrix［J］. IEEE Transactions on Geoscience and Remote Sensing， 2011， 49（6）： 2251-2258. DOI： 10.1109/TGRS.2010.2099124 .
[30]	Li D， Zhang Y H， Liang L T. A mathematical extension to the general four-component scattering power decomposition with unitary transformation of coherency matrix［J］. IEEE Transactions on Geoscience and Remote Sensing， 2020， 58（11）： 7772-7789. DOI： 10.1109/TGRS.2020.2983758 .
[31]	Chen S W， Wang X S， Sato M. Uniform polarimetric matrix rotation theory and its applications［J］. IEEE Transactions on Geoscience and Remote Sensing， 2014， 52（8）： 4756-4770. DOI： 10.1109/TGRS.2013.2284359 .
[32]	Xiang D L， Tang T， Ban Y F， et al. Unsupervised polarimetric SAR urban area classification based on model-based decomposition with cross scattering［J］. ISPRS Journal of Photogrammetry and Remote Sensing， 2016， 116： 86-100. DOI： 10.1016/j.isprsjprs.2016.03.009 .
[33]	方保镕，周继东，李医民. 矩阵论［M］. 2版. 北京：清华大学出版社， 2013.
[34]	Dey S， Bhattacharya A， Ratha D， et al. Target characterization and scattering power decomposition for full and compact polarimetric SAR data［J］. IEEE Transactions on Geoscience and Remote Sensing， 2021， 59（5）： 3981–3998. DOI：10.1109/TGRS.2020.3010840 .
[35]	王贤圆. 极化合成孔径雷达图像特征表示与目标分类方法研究［D］. 成都：电子科技大学， 2021.
[36]	Liu X， Jiao L C， Liu F. PolSF： PolSAR image dataset on San Francisco［EB/OL］.arXiv 2019： 1912.07259. （2019-12-16）［2023-12-20］. .