基于无人机高光谱遥感在太湖蓝藻水华监测中的一次应用

An Application of UAV-based Hyperspectral Remote Sensing in Monitoring Cyanobacterial Bloom in Taihu Lake

采用机载高光谱视频相机,在4个季节对太湖蓝藻进行7次、18个架次的有效拍摄。对拍摄到的高光谱影像进行辐射定标、几何拼接等预处理后,提取不同浓度蓝藻和水草等其他物体的高光谱数据,发现不同浓度的蓝藻光谱在680 nm后表现出较大差异。采用主成分分析(PCA)对高光谱数据降维后,结合k-近邻(k NN)分类算法,可实现对蓝藻的精准定位。定性识别结果经光谱预处理后,采用连续投影算法(SPA)进行特征波段提取,发现蓝藻光谱的季节差异主要表现在450 nm~570 nm和760 nm~910 nm波段。

An airborne hyperspectral video camera was used to take effective photos of cyanobacteria in Taihu Lake for 7 times and 18 sorties in four seasons.After preprocessing the hyperspectral images by radiometric calibration and geometric splicing,and extracting the hyperspectral data of cyanobacteria,aquatic plants and other objects of different concentrations,it was discovered that the spectra of cyanobacteria with different concentrations showed great differences after 680 nm.After dimensionality reduction of hyperspectral data by principal component analysis(PCA),combined with k-nearest neighbor(k NN)classification algorithm,accurate positioning of cyanobacteria could be realized.After spectral preprocessing the qualitative identification results,and extracting the characteristic bands by successive projections algorithm(SPA),it was found that the seasonal differences of cyanobacteria spectra were mainly at 450 nm~570 nm and 760 nm~910 nm.