| 高光谱遥感图像联合加权随机分类器的设计与应用 |
| |
| 引用本文： | 周前祥,敬忠良.高光谱遥感图像联合加权随机分类器的设计与应用[J].测绘学报,2004,33(3):254-257. |
| |
| 作者姓名： | 周前祥 敬忠良 |
| |
| 作者单位： | 航天医学工程研究所,北京,100094;上海交通大学,电子信息学院,上海,200030 |
| |
| 基金项目： | 国家高技术研究发展计划(863计划)
,
中国博士后科学基金 |
| |
| 摘 要： | 应用随机过程理论,提出一种针对决策融合的高光谱遥感图像自适应最小距离和K-means聚类的加权联合随机分类器设计方法.在ENVI/IDL平台上对上海市某地区的OMIS图像进行分类处理,与经典的聚类法相比,从目视判断还是从定量评价分析,它可提高约10%的总体分类精度.
|
| 关 键 词： | 高光谱图像 分类决策 聚类统计 最小距离分类 |
| 文章编号： | 1001-1595(2004)03-0254-04 |
Weighted Combination Random Classifier of High Spectral Remote Sensing Image:Design and Application |
| |
| ZHOU Qian-xiang,JING Zhong-liang.Weighted Combination Random Classifier of High Spectral Remote Sensing Image:Design and Application[J].Acta Geodaetica et Cartographica Sinica,2004,33(3):254-257. |
| |
| Authors: | ZHOU Qian-xiang JING Zhong-liang |
| |
| Institution: | ZHOU Qian-xiang~1,JING Zhong-liang~2 |
| |
| Abstract: | Based on random theory, the weighted combination random classifier of high spectral remote sensing image has been put forward. It is composed of K-means and minimum distance classification. And now it is applied to OMIS image classification of some area in Shanghai. On the platform of ENVI/IDL, the simulation results show that not only by viewing but also quantity evaluation, the total classification precision will be improved at the rate of 10% with this classification method, compared with other methods, such as maximum likelihood etc. |
| |
| Keywords: | high spectral image decision classification class statistic minimum distance classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
