Information Technology Reference
In-Depth Information
sensory modules occurs on the first network level. The experimental studies of the sensory
pattern recognition are fulfilled using “CDOT” for the presented light-colour technology of
the soil control underlying the operation of neural networks on the first level. A complex
control parameter for the technological production process of an agricultural field is formed
on the second level. The neural network on the third level enables to predict the value in a
spot of the field based on generalized parameter changes to a point of time when the
processing machine with its actuator is located at this one. To get reference colour patterns,
a special palette is developed composed of 10×10 colour cells and primary polygraphic
colours of the standard CMYK (C - cyan, M - magenta, Y - yellow, K - black) system are
presented in corner palette cells, but all the other colour tones of ones can be got by primary
colour mixing. Advantages of the used model of reference colour patterns consists in the
precise identification of palette colours and soil colour tones, respectively, but also in the
application for matching colours, e.g., Pantone (R). Surfaces of reflection coefficients for
every colour of the optical radiation are produced using the developed palette (Fig. 8) [1, 5].
The minimum Euclidian distance is chosen as a decision rule for the nearest reference
pattern (soil colour) in accordance with soil reflection coefficients registered by the sensory
system “ISSE”, but soil evaluation information is stored in the database of the intelligent
system “CDOT”.
Fig. 8. Sensory modules and neural networks (NN) in precision agriculture with
dependences of reflection coefficients for different wavelengths on the reference colour:
1 - dark-grey soil sample; 2 - light-grey one
2.4 Electronic virtual maps in precision agriculture
Having generated soil light-colour information patterns, the intelligent microsensory system
“CDOT” can produce, e.g., electronic virtual maps of the fertility level of soil spots in some
