论文部分内容阅读
采用差分偏振光谱法在3～5μm波段对水面溢油污染物进行被动遥感时,所测差分偏振光谱是含有强大气吸收光谱信号与油污染物目标弱光谱信号的混合谱,这给油污染物光谱特征识别带来了困难。另外,受环境因素以及油膜自身张力影响,水面油膜厚度分布以及油膜表面粗糙度在测量过程中发生变化,从而使得连续测量的差分偏振光谱中油污染物光谱信息含量存在不同。利用这一特点,基于固定点迭代的快速主成分分析算法FastPCA设计了水面溢油污染物差分偏振光谱信号预处理算法。实验结果表明,该算法可以有效地将水面油污染物目标光谱特征信息从具有强大气吸收的混合差分偏振光谱信号中提取出来,通过光谱重构得到油污染物光谱特征信号,可用于进一步的定性、定量分析。
When differential polarization spectroscopy is used for passive remote sensing of surface oil spill pollutants in the band of 3 ~ 5μm, the measured differential polarization spectrum is a mixed spectrum containing the strong gas absorption spectrum signal and the weak spectral signal of the oil pollutant target, which gives oil pollutants Spectral feature recognition has caused difficulties. In addition, due to environmental factors and the self-tension of the oil film, the distribution of the oil film thickness on the water surface and the surface roughness of the oil film change during the measurement, so that there are differences in the spectral information content of the oil pollutants in the continuously measured differential polarization spectrum. Utilizing this feature, FastPCA, a fast principal component analysis algorithm based on fixed-point iteration, was designed to preprocess the differential signal of polarization spectrum of surface oil spill pollutants. The experimental results show that the proposed algorithm can effectively extract the target spectral feature information of oil surface contaminants from the mixed differential polarization spectrum signal with strong gas absorption and obtain the spectral feature signal of oil contaminants through spectral reconstruction, which can be used for further qualitative , Quantitative analysis.