Definiens Centre of Excellence
As part of the Definiens Centre of Excellence, IGES has utilised the software for both research and teaching and has been involved in running Definiens short courses.
Research Using Defines Software
Mapping of semi-natural habitats and agricultural land in Wales
In 2005, the British National Space Centre (BNSC) under its GIFTSS program provided funding to the Countryside Council for Wales (CCW) to investigate the potential of satellite-based techniques for discriminating and mapping semi-naturalhabitats. The project, which was undertaken by Aberystwyth University and Environment Systems Aberystwyth and focused on the Berywn Mountains in mid north Wales (including the Lake Vrynwy area), developed a system based on Definiens object-based segmentation and classification software (Definiens,2009) for classifying semi-natural habitats (Lucas et al., 2006).
Following on from this success, the approach has been applied across Wales with a view to revising the existing national Phase 1 Habitat Map. For segmenting the landscape, SPOT-5 High Resolution Geometric (HRG) imagery area used as these provide sufficient detail to detect many features in the landscape, including hedgerows, gorse patches and wet flushes. Other imagery from optical sensors including ASTER and IRS are also included with ancillary information (e.g., digital elevation models) in a rule-base that captures key elements of the landscape that ecologists exploit to describe vegetation communities as part of their field surveys (e.g., seasonal phenology, moisture, proportion of photosynthetic and senescent material). The rule-based covers habitats extending from the coast to the uplands and includes coastal dune systems, woodlands, grasslands and heaths. The research received the Definiens GMES Innovation and Research Award in 2007. A launch took place in 2010.
Tree crown delineation and species discrimination
In 2000, fine (< 1 m) spatial resolution airborne Compact Airborne Spectrographic Imager (CASI) data were acquired with other airborne datasets (e.g., LiDAR, radar) over mixed species woodlands and open forests west of Injune in central south-east Queensland, Australia. Whilst individual and clusters of tree crowns were identifiable within the imagery, delineation proved difficult because of the variability in crown sizes associated with trees of different species composition and growth stage. However, using Definiens Developer software, an algorithm was developed that facilitated delineation of tree crowns across a range of tree sizes and architectures. Using spectra extracted from different parts of delineated crowns (Lucas et al., 2008a), maps of tree species/genus were generated (Figure 1).
The algorithm has been applied consistently across the study area and has also been successfully applied to imagery from other sensors (digital photography, LiDAR) and acquired in different regions (Europe, Chile, Canada). The algorithm has also allowed the development of techniques for biomass estimation (Lucas et al., 2008b), facilitated better parameterisation of 3D radar image simulation models (Lucas et al., 2006) and allowed the impacts of drought on tree mortality to be highlighted (Figure 1).
Figure 2. Tree crowns of some species delineated (using Definiens Developer) within 1 m CASI data in 2000 had suffered dieback because of extensive drought in Queensland.
Mangrove and wetlands characterisation, mapping and monitoring
As part of the Japanese Space Exploration Agency's (JAXA) Kyoto and Carbon (K&C) Initiative, IGES has been developing approaches to characterising mangrove structural types and detecting change using combinations of Advanced Land Observing Satellite (ALOS) Phased Arrayed L-band Synthetic Aperture Radar (PALSAR), Shuttle Radar Topographic Mission (SRTM) and Landsat sensor data. Definiens Developer software have been used to enhance the classification of mangrove type and change in several regions with extensive tracts of mangroves (e.g., Australia, south and central America). The object-orientated and rule-based approach to classification has also allowed better characterisation of wetlands.
Masters and Ph.D. Training
As part of the Masters in Geographical Information Systems (GIS) and Remote Sensing, and through Ph.D. programs, Definiens software is routinely taught to students and often forms a componet of their dissertation work. Students have access to a dedicated research laboratory with workstations running Definiens Developer software.
- Bunting, P. and Lucas, R.M. (2006). The delineation of tree crowns within CASI data of Australian mixed species woodlands, Remote Sensing of Environment, 101, 230-248
- Lucas, R.M., Lee, A.C. and Bunting, P.J. (2008). Retrieving forest biomass through integration of CASI and LiDAR data. International Journal of Remote Sensing, 29(5), 1553-1577
- Lucas, R.M., Bunting, P., Paterson, M. and Chisholm, M. (2008). Classification of Australian Forest Communities Using Aerial Photography, CASI and HyMap Data. Remote Sensing of Environment,112, 2088-2103.
- Lucas, R.M., Lee, A. and Williams, M. (2006). The role of LiDAR data in understanding the relationship between forest structure and SAR imagery. IEEE Transactions in Geoscienceand Remote Sensing, 44(10),2736-2754.
- Lucas, R.M., Rowlands, A., Keyworth, S., Brown, A. and Bunting, P. (2006). Mapping habitats and agricultural land covers in Wales using multi-temporal remote sensing data. International Journal of Photogrammetry and Remote Sensing, 62(3), 165-185.