Proximal Monitoring of CO2 Dynamics in Indoor Smart Farming: A Deep Learning and Image-Sensor Fusion Approach
Open Access
- 3 December 2025
- journal article
- research article
- Published by MDPI AG in Sustainability
- Vol. 17 (23) , 10838
Abstract
In controlled environment agriculture (CEA), CO2 enrichment can promote photosynthesis while simultaneously reducing evapotranspiration, but the optimal settings vary depending on crop type, growth stage, and microclimate. This study presents a near-field remote sensing framework that fuses RGB image features with environmental variables to predict the CO2 uptake/respiration dynamics of five leafy vegetables grown in a hydroponic culture system and evaluate their impact on resource efficiency under CO2 control. A hybrid deep model incorporating You Only Look Once version 11 (YOLOv11) and a Residual Network with 50 layers (ResNet50) extracts growth-related visual cues and integrates them with tabular features (CO2, temperature, and light conditions) to predict chamber CO2 dynamics. Performance was evaluated by Mean Absolute Error (MAE)/Mean Squared Error (MSE) on withheld data, and the system-level impacts on water use (ET), pumping energy, and relative yield were analyzed using a conventional greenhouse model. The model exhibited high accuracy (MAE = 0.95; MSE = 1.62). Scenario analysis results showed that increasing ambient CO2 concentration from 400 to 1200 ppm reduced modeled water demand by approximately 11%, increased modeled yield by approximately 9%, and resulted in a corresponding reduction in pumping energy per unit area. Unlike conventional single-crop, table-based approaches, this study demonstrates multi-crop generalization and image-environment fusion for CO2 dynamic prediction, establishing proximity sensing as a viable decision-making layer for CEA. While yield/ET results were simulated rather than measured in long-term trials, and leaf area normalization was not available, the proposed framework provides a viable path for data-driven CO2 control in indoor farms by linking image-based monitoring with operational optimization.Keywords
Funding Information
- Ministry of Trade, Industry and Energy (RS-2024-00424595)
- National Research Foundation (RS-2025-00523335)
- Korea Maritime and Ocean University Research Fund
This publication has 19 references indexed in Scilit:
- Characterization of groundwater drought and understanding of climatic impact on groundwater resources in KoreaJournal of Hydrology, 2024
- Photosynthesis prediction and light spectra optimization of greenhouse tomato based on response of red–blue ratioScientia Horticulturae, 2023
- Multi-objective optimal regulation model and system based on whole plant photosynthesis and light use efficiency of lettuceComputers and Electronics in Agriculture, 2023
- A canopy photosynthesis model based on a highly generalizable artificial neural network incorporated with a mechanistic understanding of single-leaf photosynthesisAgricultural and Forest Meteorology, 2022
- An improved whale optimization algorithm based on multilevel threshold image segmentation using the Otsu methodEngineering Applications of Artificial Intelligence, 2022
- Real-time burn depth assessment using artificial networks: a large-scale, multicentre studyBurns, 2020
- Strategies for mitigation of climate change: a reviewEnvironmental Chemistry Letters, 2020
- Modeling the canopy photosynthetic rate of romaine lettuce (Lactuca sativa L.) grown in a plant factory at varying CO2 concentrations and growth stagesHorticulture, Environment, and Biotechnology, 2016
- Comparison of three evapotranspiration models for a greenhouse cooling strategy with natural ventilation and variable high pressure foggingScientia Horticulturae, 2012
- Review: CO2 enrichment in greenhouses. Crop responsesScientia Horticulturae, 1987