Abstract: For hyperspectral data, due to the narrow sensor operating band, small sensing energy, and the various bands in the imaging process to produce the sensor radiation error and atmospheric effects are intertwined, resulting in unilateral consideration of the atmospheric factors of the correction method is difficult to obtain high-precision results. Therefore, from the principle of radiative transfer, this paper takes the Zhuhai-1 hyperspectral data as an example, combines two typical features of high and low reflectance, proposes an integrated radiation and atmosphere correction model for hyperspectral data, and compares its correction results with FLAASH, QUAC and EMPL methods, and at the same time, selects three types of typical features, namely, bare soil, vegetation and water bodies, for accuracy analysis. The results show that the correction results in this paper can effectively correct the effect of atmospheric scattering, the correlation coefficients of the correction results are all above 0.9, the spectral angle SAM are all located within 13°, the maximum root mean square error RMSE is not more than 0.15, and the correction results are stable, especially in the case of low-reflective water bodies, and the effect is much better than other atmospheric correction methods.

Key words: atmospheric correction, hyperspectral data, OHS/Zhuhai-1, integrated radiation and atmosphere correction