Processing aerial survey data using open source GIS software

John A Stevenson, Neil Mitchell, Harry Pinkerton

Outline

Nesjahraun Data source: ARSF GRASS Higher level products

Digital Elevation Model Multispectral infrared Orthophotos Field data maps Google Earth map

Conclusions

Lava flow emplaced 1.8 kyr Flowed into Thingvallavatn and continued 1 km underwater Interesting flow features: aa and pahoehoe; tension cracks,

platy-ridge zone, rootless cones

Nesjahraun, ingvellir, Iceland

Figure prepared using GMT

http://arsf.nerc.ac.uk

Sensors 39 megapixel digital camera Leica ALS50 LiDAR 11-channel Airborne Thematic Mapper (Daedalus ATM)

Data source: NERC ARSFAirborne Research and Survey Facility

http://www.osgeo.org/grass

Processing: GRASS GISFully-featured GIS

Raster Vector NVIZ

Compatible with many formatsModular structure

r.shaded.relief v.digit

GUI and CLILinux, Mac or WindowsScripts allow batch processing

Data 13 text files containing 70 million

points Time, position, intensity (for first

and last returns)Processing

Initial binning of data onto 10 m grid

r.in.xyz

From point cloud to DEM

Higher resolution DEMsLoad individual points

v.in.ascii r.in.xyz

Additional lidar processing v.lidar.*

Surface interpolation (1 m grid) v.surf.rst r.surf.rst

Problems with misregistration of lines Use individual lines 2 m grid spacing kriging via gstathttp://www.gstat.org

DEM-derived products

Shaded relief maps show surface textures: r.shaded.relief Slope maps highlight edges such as flow margins: r.slope.aspect Sun's denoising algorithm improves clarity on slope maps

http://www.cs.cf.ac.uk/meshfiltering

Data require rectification from flight-line projection Import as 11-band geotiff files

r.in.gdal 11 band images can be recombined in different combinations

r.composite red=... green=... blue=... Batch-processing saves huge amounts of time

Visible Near IR Short Wave IR Thermal IR

Multispectral infrared

An orthophoto is an aerial photo that has been warped to give a constant scale over the whole image.

Requirements to orthorectify an aerial photo:

Aerial photograph Camera geometry

parameters Ground Control Points

(GCPs) Topographic model

(DEM)

High-resolution LiDAR DEMs allow very precise orthorectification

Orthorectification

Orthorectification

Orthophotos can be used as base maps during field mapping

The photographic background allows precise navigation on complex terrain

Features such as cracks in the lava and prominent boulders are clearly visible

Fieldwork - orthophotos

gmt.soest.hawaii.edu

Fieldwork - orthophotos

Incorporating field dataGPS data imported via GPS Babel:http://www.gpsbabel.org/

KML export also available

Clickable map of photo locations via GRASS HTMLMAP driver

Geospatial Data Abstraction Libraryhttp://www.gdal.org

gdal_translate: Convert between different GIS raster file formats

Geotiff IMG NetCDF (GMT) ESRII Grid ASCII

ogr2ogr: Convert between different GIS vector formats

Shapefile KML

gdal_warp: Reproject data into different projections (uses EPSG codes) gdal2tiles.py: Export raster files for viewing in Google Earth

Windows binaries available as part of FWTools: http://fwtools.maptools.org/

Export to Google Earth (GDAL)

Conclusions Open source software is a cost-effective way to

process NERC ARSF data GRASS GIS provides most of the necessary tools Integration with other software extends

capabilities Scripts and batch processing hugely speed up

repetitive tasks

As of November 2009 I will be available for subcontracting and / or training. Contact me on: johnalexanderstevenson@yahoo.co.uk

Slide 1OutlineNesjahraun, ingvellir, IcelandData source: NERC ARSFProcessing: GRASS GISFrom point cloud to DEMHigher resolution DEMsDEM-derived productsMultispectral infraredOrthorectificationSlide 11Fieldwork - orthophotosSlide 13Incorporating field dataExport to Google Earth (GDAL)Conclusions