A new method to correct hyperspectral line scanner image data in airborne remote sensing for bidirectional reflectance effects is presented. Those effects prevent a precise intra- and intercomparison of image scenes and affect spectral ratios. The method bases on the linear semiempirical Ambrals model (Algorithm for MODIS Bidirectional Reflectance Anisotropy of the Land Surface, Lucht 2000). The samples for the inversion of the model are retrieved from the column averages which are calculated either over all pixels or separately over the pixels of each class of a spectral classification. The preclassification is supposed to lower the standard deviation within each column means in order to account for the different angular dependence for each surface. The data from a single scene is sufficient to perform an inversion. The application of this method to different sensor types is straightforward. As an example, images from the DAISEX'99 campaign in Barrax, taken with the wide-FOV hyperspectral sensor HyMap from different flight directions and times of the day, are modeled and corrected.
No takes yet. Share an insight, caveat, or question.
U. Beisl (2002) studied this question.