A research team from Skoltech’s Space Center (SC) and the Center for Computational and Data-Intensive Science and Engineering (CDISE) has come up with an approach to predict plant biomass gain based on 2D-and 3D-images, the institute’s press office reports. Their conclusions will make it possible to bolster the efficiency of precision farming both on Earth and in space.
The scientists collected statistical data by recording the growth of plants in an artificial soilless system using a 3D-camera. The information they obtained enabled them to establish the connection between the expansion of the total surface area of all leaves with the increase in the plants’ total biomass. Then the enlargement of the leaves’ area was captured using a 2D-camera and a dynamic model of plant growth was built based on this data.
To perform the experiment, the researchers used an automatic system with artificial growth capabilities, 2D- and 3D-cameras as well as sensors to collect data on the surrounding area. This system makes use of machine learning for modeling plant growth and forecasting its dynamics. According to the authors of the study, such an approach facilitates and lowers the costs of forecasting systems.
In the future, the research team intends to develop recommendatory systems for the optimization of greenhouse facilities based on the data collected and the algorithms of machine learning. The new findings from the experiment, as well as the development of efficient methods of their analysis, will also be beneficial for developing autonomous life support systems in space and on Earth.
New technological breakthroughs in precision agriculture will open new doors to fighting hunger in developing countries, enhancing food security, and boosting the efficiency of agriculture. Precision farming often faces some issues and unsolved tasks, with optimizing the use of resources being considered the main problem. To overcome this obstacle, scientists are putting together models, which make it possible to predict growth and optimize food production.
.jpg)
.jpg)
.jpeg)
