极化SAR自适应三分量分解方法
文献类型：期刊论文
| 作者 | 蔡永俊; 张祥坤; 姜景山 |
| 刊名 | 测绘学报
 |
| 出版日期 | 2016 |
| 卷号 | 45期号:9页码:1089-1095 |
| 关键词 | 极化合成孔径雷达 极化分解 基于模型 一般的散射机制 自适应 |
| ISSN号 | 1001-1595 |
| 其他题名 | Adaptive three-component decomposition approach for polarimetric SAR data |
| 中文摘要 | 介绍了原始极化SAR三分量分解中存在的问题，如负功率和散射机制模糊，并深入分析了其改进方法中仍然存在的缺陷，提出了一种自适应的三分量分解。该分解采用了更一般化的散射模型，并首次考虑了像素中存在不同旋转角的两个面或偶次散射目标，然后利用散射Alpha角确定除体散射之外的剩余主导散射机制，使面或偶次散射得到了更充分的保持。最后，从散射模型与极化相干矩阵自适应匹配的角度出发，提出了一种对负功率进行自适应优化的措施，使得负功率像素个数大大减少，从而分解更加准确有效。试验结果表明，该分解所得结果更符合实际地物散射过程，能更好地解决基于模型的分解方法中存在的缺陷。 |
| 英文摘要 | In this paper, the problems such as negative power and scattering mechanism ambiguity in original polarimetric SAR three-component decomposition are introduced, and the remaining flaws in its improved approaches are in depth analyzed. Based on these, an adaptive three-component decomposition is proposed, and more generalized scattering models are used. Because in one pixel there may exist two odd or double bounce scattering targets with different orientation angels, the proposed method firstly considers this situation, so that the surface and double bounce scattering can be preserved more sufficiently. And then the alpha parameter is used to identify the dominant scattering except for the volume scattering. Lastly, an optimization measure to the pixels with negative power is proposed, which significantly decreases the negative power pixels count, so the decomposition will be more accurate and more valid. The results show great improvements in real scattering characteristics extraction and the flaws in model based decomposition approaches can be better resolved. © 2016, Surveying and Mapping Press. All right reserved. |
| 收录类别 | EI |
| 语种 | 中文 |
| 源URL | [http://ir.nssc.ac.cn/handle/122/5612]  |
| 专题 | 国家空间科学中心_微波遥感部 |
| 推荐引用方式 GB/T 7714 | 蔡永俊,张祥坤,姜景山. 极化SAR自适应三分量分解方法[J]. 测绘学报,2016,45(9):1089-1095. |
| APA | 蔡永俊,张祥坤,&姜景山.(2016).极化SAR自适应三分量分解方法.测绘学报,45(9),1089-1095. |
| MLA | 蔡永俊,et al."极化SAR自适应三分量分解方法".测绘学报 45.9(2016):1089-1095. |
入库方式： OAI收割
来源：国家空间科学中心
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。