基于Hyperion影像的辽东湾水体信息自动分类

通过对辽东湾地区(局部)的水体、土壤及植被等主要地物波谱特征的对比分析及海水与淡水的波谱特征的对比分析,确定了高光谱遥感数据中的水体敏感波谱区间及区别海水与淡水的特征波谱区间,提出了基于连续波谱的水体信息提取与分类的利率法及双吸收深度模型,该方法在提取水体信息时不受山地阴影的影响,提高了水体信息的提取准确度。基于此方法和模型对研究区进行了海水与淡水信息的提取与分类,结果与实际情况吻合良好,为利用高光谱遥感数据进行水体信息提取分类,实现更好的水资源管理、洪水监测、海水入侵监测以及内河水入海口对海水的影响监测提供了有效的理论方法和技术手段。

Autoclassifying Water Bodies in Liaodong Bay Based on Model of Double Absorption Depth

Method of Slope to extract water from background and Model of Double Absorption Depth to distinguish
sea water and land water were put forward, based on comparative analysis of the spectral characteristics of major
objects such as water bodies, soil and vegetation etc. and the comparative analysis of the spectral characteristics of sea water and land water. Using this method to extract water bodies from background will improve the accuracy
and can not be affected by shadow. Water bodies were extracted from background, at the same time, sea water and
land water were distinguished based on this method and model in Liaodong Bay, the result shows that it can well
accord with the fact. This paper provides a new method and technique for classification and extraction of water
bodies' information to take a better management of water resources, flood monitoring, seawater intrusion monito-
ring and the monitoring of the river's impact on seawater in estuaries.