Automation of Agricultural Data Processing Using Computer Vision and IoT Technologies: An Experimental Study

The aim of the study was to analyse and evaluate the potential of integrating the Internet of Things (IoT), drone technologies, and neural networks in agriculture for effective monitoring and optimisation of agronomic processes. The experiment was conducted using sensors to collect data on soil moisture, temperature, and acidity in the fields, as well as drones for spectral imaging, which allowed the assessment of the condition of crops such as wheat, corn, and sunflower. Using the collected data, seasonal changes in growth conditions, including fluctuations in soil moisture, air temperature, and acidity, were identified, which required prompt interventions to adjust agronomic measures, such as additional irrigation or fertiliser application. Furthermore, based on the Normalised Difference Vegetation Index and detailed processing of drone images, the number, and location of stress zones in the fields were detected, caused by plant diseases and deficiencies of important nutrients. After applying neural networks to analyse the plant images, classification accuracy for wheat reached 93.5%, and for corn, 91.8%. Comparison with traditional monitoring methods demonstrated significant advantages in accuracy and processing speed. The study also showed a 12% reduction in water consumption while maintaining or even increasing crop yields due to more precise resource management and the use of precision agronomy. The high potential of applying IoT and drone technologies in the agricultural sector was confirmed for reducing environmental impact, rational use of water and energy resources, as well as improving crop yields and the efficiency of agricultural production.