2 - Spectral smile correction for airborne imaging spectrometers

Abstract

The use of airborne imaging spectrometers has become increasingly popular in the past few years. In order to eliminate spectral distortions in hyperspectral data, regular spectral recalibration of the sensor is needed, but this is not always possible due to the high cost and the time needed. One important spectral distortion is spectral smile. Several correction methods have been developed using spectral smile indicators for spectral distortion quantification. However, works analyzing the effect of other radiometric distortions on these indicators are rarely encountered in the literature. In this study, the effects of a sensor’s illumination conditions and land cover radiance variations are examined on the spectral smile indicators. A new method based on the trend line smile correction method is proposed, which uses two criteria for spectral smile quantification instead of a radiative transfer model. The method sufficiently corrects spectral smile and radiometric distortions. The highest accuracy is achieved on images depicting homogeneous areas.