Low-cost real-time greenhouse monitoring using rgb-d imaging and deep learning
A novel system integrating RGB-D imaging and deep learning for real-time, low-cost greenhouse monitoring, enabling data-driven precision crop management. It captures 3D spatial and spectral information for plant stress detection and space occupancy calculations, with real-time processing via edge computing.
Overview
This innovative solution utilizes RGB-D imaging combined with advanced deep learning techniques to offer a low-cost, real-time monitoring system for precision greenhouse crop management. By capturing both the visible spectrum and 3D spatial information, the system provides comprehensive insights into plant stress, morphology, and space occupancy. The integration of edge computing ensures real-time data processing and timely anomaly alerts, supporting effective crop management and decision-making.
Technical specifications
Core Components:
- RGB-D Imaging: Utilizes Luxonis OAK-D Pro PoE cameras to capture detailed spectral and spatial data.
- Deep Learning Techniques: Employs self-supervised learning and Masked Autoencoder for effective feature representation and semantic segmentation.
- Edge Computing: Raspberry Pi 5 devices enable real-time processing without the need for extensive cloud resources.
- Image Processing Pipeline: Quantifies plant status through features like canopy coloration, texture, area, and height.
The system is designed to detect abiotic stresses, such as interrupted irrigation or nutrient deficiencies, and calculate space occupancy efficiently.
Technology readiness level
Currently at Technology Readiness Level 5, the system has undergone initial integration and validation in controlled environments. Further validation is planned at Bayer facilities to assess commercial viability and scalability.
About University of Delaware
The University of Delaware is a comprehensive public research university and the state’s flagship, recognized for cross-disciplinary collaboration with industry. A research and technology park adjacent to campus co-locates corporate R&D with university labs and startups, with shared facilities and pilot-scale capabilities; integration with a regional health system enables clinical translation. Its Mid-Atlantic location offers quick access to talent, transportation, and nearby industrial clusters, while a statewide extension network supports testing and adoption. Research is supported by competitive federal funding from agencies such as NSF, NIH, DOE, USDA, and NASA. A dedicated technology transfer office streamlines IP, licensing, and startup formation, and corporate engagement provides a single point of entry.
delivered to your inbox.