Introduction to RSGISLib

Introduction

The Remote Sensing and GIS Software Library RSGISLib is primarily developed and supported by Pete Bunting and Dan Clewley; it was initially developed to provide the functionality we required for our research where it wasn't available in existing software packages. However, RSGISLib has evolved into a set of Python modules providing a wide range of functionality which is scriptable, creating a flexible system for data analysis and batch processing. The available modules are:

  • Image Calibration
  • Classification
  • Elevation
  • Image Calculations
  • Image Filtering
  • Image Morphology
  • Image Registration
  • Image Utilities
  • Histogram Cube
  • Raster GIS
  • Image Segmentation
  • Tools
  • Vector Utilities
  • Zonal Statistics

For installation of the RSGISLib software and tools please see the Software Page.

Other Software Tools

Alongside RSGISLib we use other tools, such as the GDAL software, KEA file format and TuiView image viewer.

GDAL

The Geospatial Data Abstraction Library (GDAL) software provides functionality to convert between many image and vector file format but it has been extended to provide a set of utilities for processing image data. The most useful utilities are:

  • gdalinfo -- report information about a file.
  • gdal_translate -- Copy a raster file, with control of output format.
  • gdalwarp -- Warp an image into a new coordinate system.
  • gdalbuildvrt -- Build a Virtual Raster (VRT) from a list of datasets.
  • gdal_merge.py -- Build a quick mosaic from a set of images.
  • gdal_rasterize -- Rasterise vectors into raster file.
  • gdal_polygonize.py -- Generate polygons from raster.
  • ogr2ogr -- Convert between different vector formats

The KEA File Format

The KEA file format developed by \citet{BuntingGillingham2013} and named after the New Zealand bird (Kea) is a HDF5 based image file format with a GDAL driver. Therefore, the format can be used in any software using GDAL, provided the KEA library is available. It offers support for large raster attribute tables and uses zlib based compression to provide small file sizes. The development was funded and supported by Landcare Research, New Zealand.

TuiView

TuiView is an open source viewer for remote sensing data, named after the New Zealand bird (Tui). Although primarily for viewing raster data but it will also display vectors. One of the main advantages of TuiView is it has a lot of functionality for viewing and manipulating Raster Attribute Tables (RAT), which the object-based classification within RSGISLib is build on top of. TuiView is cross platform and provides a consistent interface across Windows, Linux and Mac platforms. It is capable of handling very large datasets, providing overviews are generated.

Simple Script: Calculate NDVI

Within RSGISLib the user interface is Python functions. For example, the following script calculates the Normalised Difference Vegetation Index (NDVI):

$$ NDVI = \frac{NIR-RED}{NIR+RED} $$
import rsgislib
import rsgislib.imagecalc

rsgislib.imagecalc.imageBandMath("landsat8_img.kea",  "landsat8_ndvi.kea",  "(b5-b4)/(b5+b4)",  "KEA",  rsgislib.TYPE_32FLOAT)
rsgislib.imageutils.popImageStats("landsat8_ndvi.kea", usenodataval=False, nodataval=0, calcpyramids=True)

If the input image has a no data value of 0 then the script can be editted, as follows, so the output image will have a no data value of 9999.

import rsgislib
import rsgislib.imagecalc

rsgislib.imagecalc.imageBandMath("landsat8_img.kea",  "landsat8_ndvi.kea",  "(b4==0)||(b5==0)?9999:(b5-b4)/(b5+b4)",  "KEA",  rsgislib.TYPE_32FLOAT)
rsgislib.imageutils.popImageStats("landsat8_ndvi.kea", usenodataval=True, nodataval=9999, calcpyramids=True)

Here, a test is done on the NIR (b5) and Red (b4) bands to see if the value is 0. If either are zero then an output pixel value is set at 9999.

Software

Repostories