从红外高光谱资料的特点和应用现状出发，通过用晴空时观测光谱和背景光谱偏差矢量最小原理研究了特定云状下不同云量、云高和云水含量对观测光谱的影响，提出了一种新的红外高光谱资料云检测方法。从云污染视场中检测出不受云影响的通道，并用通过辐射传输模式(Radiative Transfer for (A)TOVS, RTTOV)模拟的大气红外探测器(Atmospheric Infrared Sounder, AIRS)资料和实测数据进行了方法可行性和有效性验证。结果表明，该方法能有效地提高云污染区域红外高光谱资料的利用率，可为有云覆盖情况下的大气参数反演提供有效途径。

Starting from the characteristics and application status of infrared hyper-spectral data, the influences of different cloud amount, cloud height and water content on the observed spectra under different cloudy conditions are studied according to the minimum vector deviation theory of the observed spectra and background spectra under clear sky conditions. Then, a new method for detecting cloud in infrared hyper-spectral data is proposed. The channels which are not affected by cloud are detected in the cloud-polluted field of view. The feasibility and validity of the method are verified by using both the AIRS’s data simulated by a RTTOV model and the measured data. The result shows that this method can effectively improve the utilization of infrared hyper-spectral data in cloud-polluted areas and can provide an effective approach to the inversion of atmospheric parameters under cloudy conditions.