Open Cut » Environment
Mine sites are required to rehabilitate their final landforms with a safe, stable and sustainable land cover. Traditionally, sites have been assessed from field measurements along transects, but these measurements are not necessarily representative of the whole rehabilitation site. Alternatively, information derived from Unmanned Aerial Vehicles (UAV) data can be used to map rehabilitation success and provide evidence of site suitability to support the progressive rehabilitation requirements. UAV based sensors have the ability to collect information on rehabilitation sites with full spatial coverage in a repeatable, flexible and cost-effective manner. The objectives of this research were to automatically map indicators of safety, stability and sustainability of rehabilitation using optical UAV data and object-based image analysis; and to convert these indicators into a category of polygon status based on a number of criteria. These indicators relate to erosion, vegetation composition and structure and for this case study include: mapping tall trees (Eucalyptus species); vegetation extent; senescent vegetation; extent of bare ground; and steep slopes. The eCognition Developer software was used for the object-based image analysis, which included the following main steps: (1) band stretching; (2) creating object-based canopy height model; (3) mapping vegetation extent; (4) assigning vegetation to height classes; (5) mapping extent of bare ground; (6) mapping bare ground areas with steep slopes; and (7) mapping senescent vegetation. Converting these land-cover indicators into polygon categories was used to indicate the level of rehabilitation success and these varied across sites and age of the rehabilitation that was assessed. This work presents an overall framework and workflow for undertaking a UAV based assessment of safety, stability and sustainability of mine rehabilitation and provides a set of recommendations for future work.