Information Technology Reference
In-Depth Information
Fig. 6. Correlation of the organics content in soil and its reflection of the optical radiation: (a)
well structured soil; (b) soil with a high content of sand
using the self-learning intelligent system “ISSE” it is possible to determine the content
of phosphorus and potassium in soil which is varied directly as the reflection
coefficient, but the verification of an estimated model with experimental data of
network outputs shows a high linear dependence.
The calculation of predictive models and special developed evaluation indicators in
accordance with indexes of a soil physical state was used for the recognition of soil
information patterns and for the comparison of ones with reference patterns in precision
agriculture. Then algorithms of neural networks with the genetic optimization used by us
enable to detect a set of basis information patterns of soil. These ones characterize not only
the soil individual state (Fig. 7), but also its agrophysical state in general, increasing the level
of crop yield, the quality and the biosafety of raising crops, foods, and a soil information-
microbial state.
Fig. 7. Soil information patterns using “CDOT”
Sensory information processing and the control of agricultural operations in the intelligent
system “CDOT” for precision agriculture is based on the self-learning ability of expert
systems, e.g., by means of neural network modelling. Then the recognition of
multiparameter information patterns generated by the output data transformation of
