PROBABILISTIC TOPOGRAPHIC MAPS FROM RAW, FULL-WAVEFORM AIRBORNE LIDAR DATA André Jalobeanu 1 , Gil R. Gonçalves 2 FCTUC, INESC - University of Coimbra, Portugal CGE - University of Évora, Portugal AGU FM’11 EP51E AutoProbaDTM project PTDC/EIA -CCO/102669/2008, FCOMP-01-0124-FEDER -010039
18
Embed
PROBABILISTIC TOPOGRAPHIC MAPS FROM RAW, FULL …lsiit-miv.u-strasbg.fr/paseo/slides/AGU11.pdf · PROBABILISTIC TOPOGRAPHIC MAPS FROM RAW, FULL-WAVEFORM AIRBORNE LIDAR DATA André
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PROBABILISTIC TOPOGRAPHIC MAPS FROM RAW, FULL-WAVEFORM AIRBORNE
LIDAR DATA
André Jalobeanu1, Gil R. Gonçalves2
FCTUC, INESC - University of Coimbra, PortugalCGE - University of Évora, Portugal
Over 5700 points: isolated, tracks, x-sectionsDGPS, GPRS, RTK... from 1 to 5cm vertical accuracyMostly elevation control points, a few features (buildings, sports fields)
Use raw waveform data
Sample the topography at a sufficient densityApply Shannon’s sampling theorem to avoid aliasing point spacing ≈ footprint size➔ consistent change detection, ground motion, deformation...
Make sure the calibration is done right
Fly as low as the budget permits, allow sufficient overlap
Have a good reference GPS station
Acquire some control points, just in case...
Recommendationsfor accurate, consistent and useful probabilistic DEMs
Filtering and classification- Vegetation filtering and bare earth gridding (simultaneous)- Classification, segmentation of the DEM
Computational efficiency- Process large volumes of raw data, roughly 100 GB- Complexity of the gridding and filtering algorithms
Full automation- Unsupervised gridding and classification: parameter estimation- Automatic calibration (boresight) from normal flight lines
Full uncertainty map computation- Inversion of sparse matrices for variance/covariance computation