User Tools

Site Tools


Part 1

  • Introduction to vector and raster data (raster_images.pdf)
  • Introduction to open source geodata tools

    Introduction to vector and raster data

  • raster_images.pdf

Introduction to open source geodata tools

Visualization of raster and vector data

  • openev
  • QGIS
  • exercise 1a: Use of openev and ogr tool. Reprojection of vectors and feature selection
  • exercise 1b: Convert forest vector file to KML vector for visualisation in GoogleEarth

Basic geospatial raster data processing

Format conversion: gdal_translate

Check the usage of gdal_translate on website and via command line:

gdal_translate --help

In its most basic form, without any options, gdal_transform copies an input image to an output image:

gdal_translate input output

Converting image or data formats is almost as simple. Check the available raster formats using

gdalinfo --formats

An interesting format is the VRT driver is a format driver for GDAL that allows a virtual GDAL dataset to be composed from other GDAL datasets. As opposed to physical data, the virtual data are described by XML files that are linked to some physical data sets.

As an example, the following example converts a Landsat image in GTiff format into a virtual data set (VRT):

gdal_translate 19990724_L7E_IP_CW.tif 19990724_L7E_IP_CW.vrt -of VRT

You can perform operations on the virtual data, such as scaling:

gdal_translate 19990724_L7E_IP_CW.tif 19990724_L7E_IP_CW.vrt -of VRT -scale 0 255 0 100
Image (re-)projections: gdalwarp

Check the usage of gdalwarp on website and via command line:

gdalwarp --help

The Landsat image over Cornwall is in a European projection (ETRS89/LAEA: epgs:3035). To obtain the image in the British national grid (epsg:27700), we can either do it directly:

gdalwarp -of GTiff -co INTERLEAVE=BAND -co COMPRESS=LZW -t_srs epsg:27700 19990724_L7E_IP_CW.tif 19990724_L7E_IP_CW_27700_direct.tif

Or we can combine several processing steps by using the resulting virtual data set from the previous step as input for gdalwarp.

gdalwarp -of VRT -t_srs epsg:27700 19990724_L7E_IP_CW.vrt 19990724_L7E_IP_CW_27700.vrt

We can thus postpone the creation of the physical data (in GTiff format) until the very end, which saves the creation of large temporary files and data input/output (I/O).

gdal_translate -co INTERLEAVE=BAND -co COMPRESS=LZW -of GTiff 19990724_L7E_IP_CW_27700.vrt 19990724_L7E_IP_CW_27700.tif

As a default, GDAL (gdal_translate and gdalwarp) writes GTiff bands as interleaved by pixel. By using the creation options (-co) for INTERLEAVE and COMPRESS, we can change this.

Spatial and spectral image subsetting

We want to crop the Landsat image to the following bounding box:

ULX=3217600 ULY=3127850 LRX=3225650 LRY=3119900
gdal_translate -projwin 3217600 3127850 3225650 3119900 19990724_L7E_IP_CW.tif 19990724_L7E_IP_CW_crop.tif

We now want to create a quick look of the Landsat image in JPEG format, using false colors:

gdal_translate -ot Byte -scale -outsize 20% 20% -b 4 -b 3 -b 2 -of JPEG 19990724_L7E_IP_CW.tif 19990724_L7E_IP_CW.jpg
  • Vector and raster data processing
  • Displaying vector and raster data
  • Visualizing spatial data statistics

Basic geospatial vector data processing

wiki/pktools_intro.txt · Last modified: 2017/12/05 22:53 (external edit)