海岸带高光谱影像分类技术研究  

Hyperspectral Image Classfiction in Coastal Aera

作　　者：王宏勇[1] 董广军[1] 唐汗松[1] 朱朝杰[1] 

机构地区：[1]解放军信息工程大学测绘学院,河南郑州450052

出　　处：《海洋测绘》2004年第6期20-23,共4页Hydrographic Surveying and Charting

基　　金：国家"863"计划项目(13主题)。

摘　　要：以海岸带地物分类研究为重要对象,针对传统海岸带区域地物及海水等目标分类技术的局限性,通过分析海岸带地物的光谱特征以及在高光谱数据中的表征,研究高光谱成像技术在海岸带遥感应用中的有效性。实验结果表明相对传统的遥感影像分类技术,高光谱海岸带分类显得更为精细,也展示了高光谱技术在海洋遥感中的巨大应用前景。Hyperspectral remote sensing is the focus of the remote sensing research and a great deal of achievements have been made in applications for various purpose.Hyperspectral remote sensing effectively make the spectral feature and geometric charactes of objects together.In this paper,hyperspectral image classfiction in coastal aera is studied and Experiments of hyperspectral image classification are practiced by using SAM matching model.The result of classfication is better than tradition classfication method such as ISODATA.Research indicates that the diversity and complex can be distinguished by using hyperspectral data.So hyperspectral image shows great foregroud in ocean remote sensing.

关 键 词：海洋遥感 海岸带 高光谱 影像分类 

分 类 号：P237[天文地球—摄影测量与遥感]