Laserscanning aktueller Stand in Forschung und Entwicklung Karl Kraus Institute of Photogrammetry and Remote Sensing (I.P.F.), Vienna University of Technology www.ipf.tuwien.ac.at Neues gegenüber dem Vorjahr Beispiele vor allem aus dem Pulkauprojekt (N) Gebude und Spezifisches zum Hochgebirge erst am Nachmittag
30
Embed
Laserscanning Œ aktueller Stand in Forschung und … · Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 2 Geo-Referencing Observations Łheights of the control
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
Laserscanning � aktueller Stand in Forschung und Entwicklung
Karl Kraus
Institute of Photogrammetry and Remote Sensing (I.P.F.), Vienna University of Technology
www.ipf.tuwien.ac.at
� Neues gegenüber dem Vorjahr
� Beispiele vor allem aus dem Pulkauprojekt (NÖ)
� Gebäude und Spezifisches zum Hochgebirge erst am Nachmittag
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 2
Geo-Referencing
Observations�heights of the control points
�height differences of parallel plane patches calculated from neighbouring points in adjacent strips
Unknowns�heights for all the tie points
�heights for all the control points
�additional parameters for the compensation of the systematic errors
Improvement of the homogeneity in the overlapping areas by a factor between 2 and 4
=> Extension to 3D (with structure line elements) and modelling the flight path, etc.
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 3
Geo-Referencing
�Control Point� = a few tilted planes
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 4
Projekt Pulkautal
Differenzmodell der Laserscanner-Streifen 10 und 11
nach der simultanen Höheneinpassung
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 5
Derivaton of DTMsRobust Linear Prediction
271 control points:
σ = ±18cm
point distance : ~3.1m
σ = ±29cm
point distance: ~1m
=> point density is the most important parameter
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 6
Derivation of DTMs (cont.)Robust linear prediction in a hierarchical approach
a) Creation of a data pyramid (lowest point with its xyz-co-ordinates in a regular grid)
b) DTM generation in the coarse level
c) Coarse DTM with a tolerance band
d) DTM generation in the fine(r) level
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 7
Derivation of DTMs (cont.)
Laser scanner data with the gaps after the elimination of theoff-terrain points
Kamptal in NÖ, Department of WaterManagement, Hydrology and Hydraulic EngineeringBOKU - University of Natural Resources and Applied Life Sciences, Vienna)
Quality Overlays� distance between each grid point and the next point of the original data set
� variances in small cells derived from the differences between the elevation of the original points and the elevation of the DTM at the same locations
� the density of the points also in this small cells
� σ in each grid point
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 8
Projekt Pulkautal
[Punkte/m2] [Punkte/m2]
Punktdichte der original Laserscanner-Punktwolke
Punktdichte der Bodenpunkte
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 9
Projekt Pulkautal (cont.)
Abstandsmodell für die Bodenpunkte
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 10
Projekt Pulkautal (cont.)
Mittlere Residuen zwischen Laserpunkten und DGM
ursprüngliche Laserpunkte Bodenpunkte
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 11
Projekt Pulkautal (cont.)
Manuelle photogrammetrischeDatenerfassung
Photogrammetrisches DGM
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 12
Projekt Pulkautal (cont.)
Laserscanner-DGM Orthophoto
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 13
Projekt Pulkautal (cont.)
� Photogrammetrie �Modell-Patches�
� Scanlag in ALS
� Kanten in Photogrammetrie
� Kleinformen in ALS
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 14
Accuracy and Geomorphologic Quality of Laser Scanner DTMs
Empirical elevation accuracy of LaserScan DTM
tanα . . . Terrain slopen . . . Average number of points per square meter
[ ]
⋅+±= ασ tan1206n
cmz
For flood risk area the accuracy of the positionof the border of predicted flood is of central interest:
ασσ
tanz
xy =
e.g. σz = +/- 12 cm σxy = +/- 25 m ! (tan α = 5%o)
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 15
Accuracy and Geomorphologic Quality of Laser Scanner DTMs (cont.)
Flood risk area (project of the Lower-Austrian Govermental Administration)
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 16
Geomorphologic Quality of Laser DTMs� only a point cloud� point density is the important parameterWavelike surfaces with a minimal wave length Lmin
Scanning theorem
∆ ≈ Lmin/3
Derivation of breaklines from the point cloud
minimal crown width Bmin
∆ ≈ Bmin/2
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 17
Surface Break Lines
Hybrid DTM consisting of a regular raster with 3D vectors
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 18
Surface Break Lines (cont.)
How to get the relevant 2D break lines?
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 19
Surface Break Lines (cont.)
Automatic procedure: rain simulationSemiautomatic procedurea. Manual digitization as 3D lines in photogrammetic stereomodels
Almtal
Upper-Austria
Govermental
Administration
BEV: The entire territory of Austria* image scale 1:15 000, 21 cm camera* σz = +/- 45 cm (too bad)* σxy = +/- 25 cm (good enough)
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 20
Data Reduction
� The very high density of points is a great advantage for improving the quality
� but there are difficulties in the data management.
Derivation in each grid point of the dense DTM:
� radius R of the curvature
� slope tanα
� allowable distance E in order to keep dZ less than threshold dZmax
Idea of progressive sampling
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 21
Data Reduction (cont.)
Small example
Wienerwald
MA 41 der
Stadtverwaltung
Wien
Typical example
� 4 millions of grid points
� user defined threshold dZmax = 25cm
� 1.3 million of the reduced points
� 4.75 minutes on a 2 GHz computer including the output in ASCII
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 22
Monitoring Flooding Catastrophes
Monoplotting
Different types of images, geo-referenced by GPS and IMU:
� Digital frame camera
� One-line scanner
� Three-line scanner
� Amateur camera (with oblique axes)
� Microwave sensors (for extreme weather conditions in flooding periods)
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 23
Monitoring Flooding Catastrophes (cont.)
Scenario� Actual water level and the 3D border lines of the
flood, corresponding to the time when the images have been taken
� Based on short-term prognoses of precipitation,prediction of the rise in water level
� Increase of the actual water level by the predicted rise ⇒ Surface at the predicted water level
� Intersection of the surface of the predicted water level with the precision DTM
⇒ Areas at immediate risk can be plotted
⇒ Catastrophe management can take important decisions and can inform those affected.
A vision: After taking the images final results within one day
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 24
Additional Features Relevant to Hydraulics and Hydrology
Model without structure lines:- grid width: 50 m- 55.000 points
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 25
Additional Features Relevant to Hydraulics and Hydrology (cont.)
Hydrological analysis:- drainage
dark-blue: valleylight-blue: watershed
developed by:
Rieger W.
Gajski D.
Molnar L.
Dorninger P.
Karl Kraus Laserscanning - aktueller Stand in Forschung und Entwicklung 26
Additional Features Relevant to Hydraulics and Hydrology (cont.)