一种全极化高分SAR与中分光学影像融合方法
作者简介:万剑华(1966—), 男, 山东省单县人, 从事3S技术的应用研究。E-mail: wjh66310@163.com
收稿日期: 2016-06-03
要求修回日期: 2016-11-07
网络出版日期: 2017-04-06
基金资助
山东省自然科学基金项目(ZR2016DM16)
国防科技工业局高分辨率对地观测系统重大专项“海岸带遥感监测与应用示范”
中国石油化工集团科技攻关项目“航测与遥感技术在管线检测中的应用研究”
A fusion method of high-resolution full polarimetric SAR and moderate-resolution optical image
Received date: 2016-06-03
Request revised date: 2016-11-07
Online published: 2017-04-06
Supported by
Natural Science Foundation of Shandong Province of China (ZR2016DM16)
Coastal Zone Remote Sensing Monitoring and Application Demonstration of High Resolution Earth Observation System Major Projects
A Application Study on Aerial Survey and Remote Sensing Technology in Detection of Pipeline
Copyright
针对全极化SAR(synthetic aperture radar, 合成孔径雷达)与中分光学影像的融合问题, 提出基于主成分分析(PCA)与HSV(hue, saturation, value, 色调、饱和度、明度)色彩空间变换的遥感影像融合方法。对全极化SAR四个极化波段进行主成分分析, 提取第一主成分, 将中分光学影像变换到HSV空间, 第一主成分替换V分量, 用新的V分量逆变换到RGB空间, 得到全极化SAR与中分光学的融合影像。通过利用Radarsat-2全极化SAR与TM/ETM+中分光学影像开展融合实验, 结果表明, 该方法优于传统融合方法(PCA变换、HSV变换、小波变换等)的单极化SAR与光学影像融合结果, 能够有效利用全极化SAR的纹理信息, 提高影像解译能力。
万剑华 , 臧金霞 , 刘善伟 , 任广波 . 一种全极化高分SAR与中分光学影像融合方法[J]. 热带海洋学报, 2017 , 36(2) : 79 -85 . DOI: 10.11978/2016055
A method was proposed to solve the fusion problem of full polarimetric synthetic aperture radar (SAR) and moderate-resolution optical remote sensing image, based on principal components analysis (PCA) and HSV transform, where HSV stands for hue, saturation, and value. First, the four bands of full polarization SAR went through principal component analysis, and the first principal component was extracted. Second, moderate-resolution optical image was transformed into the HSV space, and the V component was replaced by the first principal component. Finally, the desired image was obtained by using inverse IHS transform. The experiment was performed using Radarsat-2 full polarimetric SAR and TM/ETM+ moderate-resolution optical image. A comparison of the proposed method with the fusion of single polarimetric SAR and optical image based on a traditional method (PCA transform, HSV transform, or wavelet transform) showed that the proposed method made effective use of the texture information of full polarimetric SAR and improved the capability of image interpretation.
Fig. 1 Technique flow chart图1 技术流程图 |
Fig. 2 The original image: a. 2010 TM image; b. 2015ETM+image; c. 2009 full polarimetric SAR false-color image; d. 2015 full polarimetric SAR false-color image; e. 2009 HV polarization SAR image; and f. 2015 HV polarization SAR image图2 原始影像图 |
Fig. 3 TM image, SAR image and fusion image in 2009 and 2010: a. 2010 TM image; b. 2009 PCA first component SAR image; c. PCA+HSV fusion image; d. PCA fusion image; and e. HSV fusion image图3 2009年和2010年TM影像、SAR影像、融合影像 |
Fig. 4 ETM+ image, SAR image and fusion image in 2015: a. 2015 ETM+ image; b. 2015 PCA first component SAR image; c. PCA+HSV fusion image; d. PCA fusion image; and e. HSV fusion image图4 2015年ETM+影像、SAR影像、融合影像 |
Tab. 1 Quantitative evaluation of the fusion results表1 融合结果定量评价表 |
遥感影像 | 均值 | 熵 | 平均梯度 | |
---|---|---|---|---|
2009、2010年遥感影像 | TM影像 | 117.891 | 5.675 | 16.686 |
HH极化 | 124.750 | 5.726 | 16.095 | |
HV极化 | 125.578 | 5.738 | 14.170 | |
VH极化 | 122.281 | 5.688 | 14.257 | |
VV极化 | 128.422 | 5.719 | 13.999 | |
PCA第一主成分 | 139.093 | 5.769 | 17.792 | |
本文方法 | 123.125 | 5.750 | 17.241 | |
PCA变换法 | 127.000 | 5.738 | 14.119 | |
HSV变换法 | 129.828 | 5.813 | 15.655 | |
2015年遥感影像 | ETM+影像 | 125.516 | 5.633 | 13.828 |
HH极化 | 125.141 | 5.813 | 16.040 | |
HV极化 | 119.656 | 5.781 | 16.787 | |
VH极化 | 123.375 | 5.844 | 15.657 | |
VV极化 | 134.936 | 5.781 | 14.783 | |
PCA第一主成分 | 136.641 | 5.856 | 16.836 | |
本文方法 | 127.672 | 5.832 | 16.166 | |
PCA变换法 | 132.156 | 5.770 | 15.164 | |
HSV变换法 | 142.016 | 5.847 | 16.208 |
The authors have declared that no competing interests exist.
1 |
|
2 |
|
3 |
|
4 |
|
5 |
|
6 |
|
7 |
|
8 |
|
9 |
|
10 |
|
11 |
|
/
〈 |
|
〉 |