Today's farming practices produce a wealth of data about the crop, the field, its soil, structure and cropping history. To make the best management decisions, you need to be able to access the right information in any place at any time. This is what precision agriculture/farming (PA) is all about, being a farming management concept based on observing, measuring and responding to inter and intra-field variability in crops.
Although remote sensing has been used extensively in the past decades to detect and monitor stress occurrence in agricultural fields, the transition of this research to real time applications useful for the farmer has been very limited. Two of the main reasons are:
- Cloud obscuration. Due to the maritime temperate climate in Belgium is the occurrence of cloud free days limited. This obstructs the timely collection of RS data, and as such severely hampers the direct applicability of RS data in a precision farming context;
- Difficult transition from the information collected with the sensor to specific task maps for the farmer.
It is only since the upcoming of the Unmanned Airborne Vehicles (UAV) that a timely, actual, accurate, detailed and objective estimation of crop growing conditions can be guaranteed by RS. UAVs however lack the spatial coverage that can be obtained by satellite sensors and the spectral range that can be obtained by the APEX sensor. Combining existing sensors and technologies with crop growth models enables us to issue yield forecasts at a range of spatial scales.
In this framework, investigation into, and development of novel methods on time series analysis, multiresolution spatiotemporal data fusion, unmixing, inverse modeling and data assimilation is performed in our group. The team also collects RS and in-situ data to support the RS analysis.
Detection of fire blight infections in pear orchards based on UAV imagery
Digital camera image Infections shown in red