光谱微分技术在高光谱遥感浅海海底底质中的应用初探

Preliminary Study on the Seafloor Type Classification by Using Spectral Derivative Method

采用光谱微分技术进行了高光谱遥感反演浅海海底底质的初步研究。首先,利用半分析海洋辐射传递模型模拟不同海底底质的光学浅水高光谱遥感反射比数据集;然后,利用光谱微分技术建立海底底质识别算法;最后利用水槽模拟数据来验证。结果表明本文建立的反演算法可以应用于泥砂类和植物类底质的区分。

The feasibility of shallow water＇s seafloor type classification with hyperspectral remote sensing based on spectral derivative technique has been investigated.Firstly,the hyperspectral remote sensing reflectance of the shallow water with different seafloor types was simulated with the semi-analytical numerical model.Secondly,the spectral derivative method was applied to the simulated data to analyze the spectral characteristics of the seafloor of different types.Finally,the water pool experiment was carried out to validate the above method.The validating results show that the seafloor covered by sand and seagrass can be detected with the method developed in this study.