| 引用本文: | 卞春江,侯晴宇,赵晓,梁冰冰,李立源,张伟.特征空间线性降维压缩遥感图像云检测方法[J].哈尔滨工业大学学报,2014,46(1):29.DOI:10.11918/j.issn.0367-6234.2014.01.006 |
| BIAN Chunjiang,HOU Qingyu,ZHAO Xiao,LIANG Bingbing,LI Liyuan,ZHANG Wei.Cloud detection in remote sensing image based on linear dimension compression[J].Journal of Harbin Institute of Technology,2014,46(1):29.DOI:10.11918/j.issn.0367-6234.2014.01.006 |
|
| 本文已被：浏览 1999次 下载 1409次 |
码上扫一扫！ |
|
|
| 特征空间线性降维压缩遥感图像云检测方法 |
|
卞春江1,2, 侯晴宇1, 赵晓3, 梁冰冰4, 李立源1, 张伟1
|
|
(1.哈尔滨工业大学 空间光学工程研究中心,150001哈尔滨;2.中国科学院国家空间科学中心, 100190 北京; 3.上海卫星工程研究所, 200090 上海; 4.哈尔滨师范大学 数学系, 150025 哈尔滨)
|
|
| 摘要: |
| 针对遥感图像云检测过程中分类特征空间维数过高引起的信息冗余,提出了一种基于特征空间线性降维压缩的云检测方法．首先选取云与地物的分类特征参量,构造特征空间,基于压缩子空间分类信息表述的完备性,建立样本的概率分布模型．然后利用最大似然估计法求解模型参数,估计最佳转换矩阵,进行特征空间的降维压缩与去相关处理．最后针对压缩子空间,利用分类器进行云检测．实验结果表明: 本方法能够有效地去除云与地物分类特征之间的冗余,实现二维压缩子空间中云与地物两类样本的有效分离,对于实际光学遥感图像的云检测概率高达98％以上． |
| 关键词: 光学遥感 云检测 特征空间 降维压缩 去相关 |
| DOI：10.11918/j.issn.0367-6234.2014.01.006 |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(61007008). |
|
| Cloud detection in remote sensing image based on linear dimension compression |
|
BIAN Chunjiang1,2, HOU Qingyu1, ZHAO Xiao3, LIANG Bingbing4, LI Liyuan1, ZHANG Wei1
|
|
(1. Research Center for Space Optical Engineering, Harbin Institute of Technology, 150001 Harbin, China; 2. National Space Science Center, Chinese Academy of Sciences, 100190 Beijing, China; 3. Shanghai Institute of Satellite Engineering, 200090 Shanghai, China; 4. Dept. of Mathematics, Harbin Normal University, 150025 Harbin, China)
|
| Abstract: |
| According to the information redundancy caused by high classification feature space dimension, we have proposed a method of cloud detection based on linear dimension compression for feature space. First, the classification feature parameters were extracted and the feature space was established. According to the different distribution structures of cloud and underlying surface samples, based on the completeness of classified information in compressed subspace, the sample’s probability distribution model is established. Then the model parameters are solved by using the maximum likelihood estimation method and the optimum transformation matrix is estimated, which is used for features compression and decorrelation. At last, a classifier is introduced for cloud detection in the compressed subspace. Experimental results have shown that the proposed method can effectively remove the classification features redundancy of the clouds and underlying surfaces, and achieve effective separation of clouds and underlying surfaces in compressed subspace. The cloud detection probability in optical remote sensing images is up to 98％ or more. |
| Key words: optical remote sensing cloud detection feature space dimension compression decorrelation |
