NEUROPUBLIC is an active member of the European Association of Remote Sensing Companies, that develops cloud EO-based services for the agricultural sector. We are doing business by using Copernicus and IoT data, promoting the penetration of EO deeper into the CAP line of business and also the establishment of the digital farming market (figure 1).
AgriMonitor is a web app that allows the monitoring of agricultural areas through time, either by using interactive graphs of vegetation and soil indices, or image snapshots that describe the intra-field crop growth conditions at various spatial and time resolutions. The information context is enriched by processed data coming from our own network of telemetric stations.
For most areas of the Greek Territory, agronomists and farmers are now able to understand at parcel level the evolution of their crops. It’s a time machine that enables decisions under a straightforward way (figure 2).
The end user may upload his agricultural parcels or he can digitize an Area of Interest (AOI) and see the results in real time. The AOI polygons may also represent large areas, such as wetlands, grasslands or even lakes and forests. This way, Agrimonitor can arise as a valuable way to monitor the evolutionary change of whole ecosystems through time.
Methodologically, at NEUROPUBLIC we have adapted an object-based approach, aiming to extract with an automated way objective and well-documented information. Each agricultural parcel, is an object under monitoring. We assign extracted indicator values to agricultural parcels, having as main input streams of Earth Observation-based layers, ground-sensor measurements, or even farm logs.
The assigned info creates a multi-dimensional signature of the object that describes the parcel under different aspects. We present the assigned info by using graphs or images. We can also use these datasets to create models that allow the extraction of insights, or the classification using various criteria. The services as well as the processing workflows have been created using open source software. We mainly use Python in the backend as there are great packages available to handle spatial data along with zonal operations and arrays. Considering the maps and the charts presented in UI, we focused on interactive libraries, such as Leaflet and Plotly. The existing infrastructure of Agrimonitor has already been tested in the development of Smart Farming and CAP services. Our future steps include the integration of Sentinel 1 and 3 data as well as the extension of the coverage at a EU level. NEUROPUBLIC will demonstrate Agrimonitor at the upcoming 23rd MARS Conference that will take place in Gormanston, Ireland, on 28 and 29 November 2017.