S. S. ATANASOV1, P. I. DASKALOV2 and V. I. NEDEVA1
1 Trakia University, Department of Electrical engineering, Electronics and Automation, BG-8600 Yambol, Bulgaria
2 University of Ruse, Department of Automatics and Mechatronics, BG-7017 Ruse, Bulgaria
ATANASOV, S. S., P. I. DASKALOV and V. I. NEDEVA, 2016. An intelligent approach of determining relationship between tomato leaves color and soil moisture and temperature. Bulg. J. Agric. Sci., 22: 1027–1035
In this article is presented a model of the relationship between the soil moisture and its temperature and the color of the leaves of greenhouse tomato plants. This dependency is modelled using nonlinear regression statistical Quasi-Newton method. Piecewise linear regression with breakpoint models are received. The impacts of the factors non-included in the model are taken into account, as they are equated to the random error. A methodology for collection, characterization and processing of information from the experiments is proposed. During the experiment wireless sensors for soil moisture and temperature, handheld portable colorimeter and combined device for measuring environmental parameters are used. In statistical processing of the information the software platform Statsoft STATISTICA is used. The obtained models predict soil moisture based on the measured RGB values and the soil temperature, with an error ranging between -4.85% to +14.98%. The resulting dependency allows creation of an intelligent system for optimal managing the process of irrigation of greenhouse plants.