Scientists leverage video imaging to advance tomato production monitoring
A team of researchers from Hebrew University and the Sami Shamoon College of Engineering, Israel, has developed a “groundbreaking” low-cost method for monitoring plant growth using standard video footage. The findings could help advance precision agriculture globally.
The technique, which reconstructs three-dimensional models of plants from 2D videos, enables accurate estimation of total leaf area (TLA) in dwarf tomato plants — without the need for expensive sensors or destructive sampling.
Led by Dmitrii Usenko, a PhD candidate at the Institute of Environmental Sciences at Hebrew University, the study applies structure-from-motion algorithms and machine learning to create detailed 3D representations of plant foliage.
The researchers used over 300 video clips of tomato plants grown in greenhouses, achieving a coefficient of determination (R²) of 0.96 with their best-performing model — outpacing traditional 2D-based approaches.
“Accurate measurement of total leaf area is crucial for understanding photosynthesis, water use, and overall plant health,” says Dr. David Helman, Usenko’s advisor and co-author.
“What’s exciting here is that we’ve combined accessibility and accuracy — bringing precision agriculture within reach for many more growers.”
Unlike conventional methods that rely on costly Light Detection and Ranging imaging or multispectral imaging, this novel approach uses only basic Red, Green, and Blue cameras. It is non-invasive, crop-independent, and scalable — opening the door for widespread use in greenhouse operations and open-field agriculture alike.
“By eliminating the high cost of monitoring tools, we’re helping democratize access to smart farming,” adds Usenko. “Our hope is to empower both smallholder farmers and large-scale operations with data-driven insights.”
The study, titled Using 3D reconstruction from image motion to predict total leaf area in dwarf tomato plants, is published in Computers and Electronics in Agriculture.
