LiDAR intensity correction

The second LiDAR intensity paper ..... here we find new information about topographic and distance effects.

Next time I write about photogrammetric and LiDAR integration opportunities. Very interesting research area!


TOPOGRAPHIC AND DISTANCE EFFECTS IN LASER SCANNER INTENSITY
CORRECTION
S. Kaasalainen, A. Vain, A. Krooks, A. Kukko

"ABSTRACT:
The effect of incidence angle on the intensity of laser backscatter has been studied in photonics and optics, but the applications of these results to remote sensing of natural land targets are limited, as well as the availability of experimental validation data for airborne and terrestrial laser scanning, where the incidence angle correction using the Lambertian scattering law is common. We have investigated the role of topographic (incidence angle) and distance effects in the radiometric calibration of monostatic terrestrial and airborne laser scanner data. Our results show that the Lambertian (cosine) correction is practically valid at incidence angles up to 20º, whereas at greater angles of incidence, the accuracy of data is still very limited to estimate the performance of any correction method. We also discuss the mixed effects of distance and target reflectance on terrestrial laser scanner intensity calibration, for which the number of applications is constantly increasing. As there are differences in the intensity detectors of different instruments, it is important that the effects of distance and target reflectance are well defined before using any terrestrial laser scanner in intensity
measurement.


SUMMARY

At the current levels of accuracy of ALS intensity data, the cosine correction of the incidence angle effect works reasonably well for most surfaces, even in the cases that the surface could not be approximated to be Lambertian. Together with earlier results (Kukko et al., 2008), it can be concluded that for most targets the incidence angle effect is practically within the error limits of the data at incidence angles up to about 20°. The computational cosine effect is about 6% in this range, which is well in the error limits of the intensity data produced by current ALS instruments (this is clearly seen in Fig. 5 and Kaasalainen et al., 2009a). It must also be taken into account that some surfaces may present a specular reflection at 0° (normal incidence), which may cause a peak in the intensity at this angle. More data with better accuracy than that provided by the current airborne scanners would be needed in cases where the incidence angle is greater than 20°, to be able to distinguish any difference between surfaces for which the Lambertian approximation does or does not work adequately. The intensity and distance effects in TLS are mixed, at least for some scanners. A further study on distance effects in TLS is in progress, with a comparison with data from other instruments (Kaasalainen et al., 2009b). This is particularly important in the applications where TLS are used in mobile (vehicle based) mapping systems and other (stationary) applications, where the distance of the target varies in a large scale. It is important that the detector effects are well known (to provide a correction scheme) before using any terrestrial laser scanner in intensity measurement."