Videographic observations can provide useful information at a scale intermediate between the large-scale data collected on the ground and the regional scale data available from satellite imagery. Previous work on agricultural, hydrologic and forestry applications report multi-spectral video as a non-invasive and rapid method for generating timely information that can be integrated with other ancillary data for better management strategies. Accordingly, this study investigates the potential of digital multi-spectral video to identify variations in crop conditions related to farm management, soil types, and terrain conditions. Specific spectral, spatial and temporal remote sensing requirements for mapping variations in crop conditions are addressed as well. Vegetation indices were applied to map different crop types and conditions. The NDVI, plant pigment ratio, plant vigour index and cell density ratio were used to this end. The study examines whether the spectral variability observed in the vegetation indices of paddocks under crops are associated with terrain attributes such as slope and aspect, different soil-landscape types, waterlogging, and crop conditions such as the presence of weeds. It is concluded that qualitative images as provided by the vegetation indices implemented in this study can provide useful information for identifying zones that perform differently within or between paddocks. The indices showed sensitive to variations in drainage conditions, soil-landscape units (e.g. and associated terrain attributes such as slope) and the presence of weeds within a paddock. Thus, it is concluded that rapid mapping of the occurrence of field variations by applying vegetation indices derived from high resolution airborne videography would enable farmers to identify the causes of variability (e.g. waterlogging, weeds, insufficient fertilisers, etc.), helping to decide on appropriate management practices for improving farming conditions.
展开▼
