System development and processing techniques for the estimation of plant's physiological parameters

Considering the expected increase in world population over the next decades, the demand for food is bound to rise as well which implies the necessity for new and scientific driven approaches in the agriculture sector. The latter is particularly suffering from global warming by fertile land becoming more and more desertic, less availability of water, changes in seasonality patterns and so on. This situation justifies the necessity to carefully manage resources for cultivation which implies the fact that we must know the needs of the plant, i.e to monitor its status and environment as to provide what it is only needed for its growth. This thesis focuses on a new approach to indirectly evaluate the physiological parameters of plants by measuring their impedances through a pair of 4 needles inserted into the plants’ stem. The plants under analysis are an apple tree and a tomato plant. The first is located in an apple orchard while the second in a climatic chamber. Various techniques are scouted for extrapolating information from the collected signals, in particular the focus is in quantifying the linear correlation between plant’s impedances and its physiological parameters via MLR, cross-correlation, Pearson correlation and Fourier analysis. The results show a good degree of linearity, especially if lag values are taken into account.