OMIS成像光谱数据信噪比的估算

实用型模块化成像光谱仪 (OperationalModularImagingSpectrometerOMIS)是一种带有稳定平台的新型机载遥感器。具有 12 8波段 ,覆盖了从可见光到热红外的光谱范围 ,其应用范围有地质、农业、林业和海洋等领域。在分析OMIS数据中的噪声特性之后 ,指出OMIS系统中各通道的噪声是和信号无关的加性噪声 ,噪声自身并不相关。并且对比分析了现有的 4种信噪比估算方法。指出基于空间和光谱维去相关的方法比较可靠 ,并给出实际的计算结果。最后指出在大气订正后 ,图像数据的信噪比能充分反映仪器的信噪比性能。

Estimation of Signal-Noise-Ratio from Data Acquired with OMIS

Operational Modular Imaging Spectrometer (OMIS) is a new type of airbome remote sensor. It has 128 spectral bands ranging from visibl e to thermal infrared wavelength. Its application fields include geography, agri culture, forest and ocean etc. For the convenience of the wide use of OMIS Data and the instrument improvement, quantitative evaluation of OMIS data is indispen sable. Signal-Noise-Ratio( SNR ) is an important evaluation scale. According to R. E. Roger's view, for a complex instrument such as AVIRIS, it is not reasonabl e to assume that the noise included is random and not correlative, especially in visible spectral bands. In this paper, through the analysis of the noise charac ters in OMIS data the authors point out that the noise mixed in OMIS data is a k ind of additive noise and not correlative. Based on athrough analysis of the fou r available SNR estimation methods, a conclusion can be drawn that the spati al a nd spectral decorrelation method is the most reasonable one for OMIS data. The e xperimental results are also presented. It is also pointed out that after radiom etric correction, the SNR of the acquired data can effeciently reflect the SNR character of the remote sensor.