Within the framework of the project supported by the NKFI (National Research, Development and Innovation Office) Fund, ULYSSYS Kft. and Komunáldata Kft. conducted joint research and development in the field of the agricultural use of satellite remote sensing.

Within the framework of the project supported by the NKFI Fund, ULYSSYS Kft. as consortium leader and Komunáldata Kft. as consortium member conducted a joint R&D project aimed at developing an innovative crop yield monitoring solution based on the processing of space images. The main objective was to provide modern, precision crop production with the most accurate data possible, for which the remote sensing technology used provided an effective tool.

Economic and social processes require agricultural activity to maximise the benefits and minimise the environmental impact. Precision crop production, which has been expanding over the last 25 years, can provide a number of appropriate responses to this challenge. This is because in precision farming, each small unit of cultivated land (the management zone), which can be considered homogeneous from a farming point of view, is treated according to its characteristics and the needs of the crop it produces. One of the most important input data for precision farming is the exact value of crop yields per cell (about 0,1 ha), which are the building blocks of the zones. Our development provides an innovative solution for this precision yield measurement.

Our yield measurement method is based on the processing of satellite imagery, exploiting the characteristics of the radiation values reflected by different crop areas at different wavelength ranges and their model correlation with plant photosynthesis. The initial data are multispectral satellite images of the area from seedbed preparation to harvest. As a result of the research carried out in this project, we can estimate the actual yield of each zone with an accuracy of 93-95%.

The main task of our research was the transformation of the basic model previously developed for the field scale, since precision farming requires data at a much smaller spatial scale than the field scale, and models that work well at the full field scale cannot be mechanically converted to the scale of zones and cells.

Therefore, in the applied research we have conducted, we have performed the transformation of our methodology using data from representative sample areas, using satellite imagery, and we have performed the experimental developments that will put our solution on a state-of-the-art SaaS (software-as-a-service) service platform.

A significant advantage is that our method can be used to measure yields retrospectively, allowing immediate entry into precision farming (as opposed to the multi-year trial period of alternative methods).

The application of the implemented yield measurement solution goes beyond serving precision farming, it can be used to objectively and credibly serve the information needs of all market and government actors for whom the accurate determination of land yields is of importance (e.g. insurance companies, banks, authorities).