smart-farming
Precision agriculture is a management approach that gathers and analyzes temporal and spatial data about plants and environmental conditions to optimize decision-making and resource use. For plant science, this data-driven methodology enables researchers to understand plant responses to varying environmental conditions at individual and population scales, improving both research precision and agricultural productivity. By systematically monitoring and responding to plant variability, precision agriculture advances knowledge of plant physiology, stress responses, and optimal growing conditions that would be difficult to discern through traditional methods.
open_in_new WikipediaPubMed · 2026-01-01
Researchers developed an artificial intelligence system that uses photographs of soil surface to predict moisture levels at different depths, enabling farmers to customize watering for individual plants rather than treating entire fields uniformly. This non-invasive approach could significantly improve crop yields and reduce water waste by adapting irrigation to each plant's specific physiological needs.
DenseNet121 deep learning model achieved 97.3% accuracy (R²) predicting surface soil moisture from RGB images with RMSE of 4.14
Random forest regression model achieved 90.6% accuracy for deeper soil layers (10-15 cm) with RMSE of 4.97, effectively capturing nonlinear moisture dynamics
Image-based analysis provides a non-invasive, scalable alternative to traditional soil sensors for real-time plant-level moisture monitoring
