Location: Cotton Ginning Research
Project Number: 6066-30600-001-001-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Sep 29, 2026
End Date: Sep 28, 2027
Objective:
The overarching goal of this project is to enhance the efficiency, accuracy, and interoperability of industrial hemp research through the development of advanced phenotyping technologies and integrated digital data systems. Achieving the goals of this project will aid in the development of a new domestic industrial hemp industry which will provide opportunities for American farmers to pursue a new crop to help with long-term soil health and increased profitability, in keeping with the research priorities of the Secretary of Agriculture.
Objective 1 is to develop a scalable, non-proprietary computational pipeline for extracting key morphological traits from high-resolution three-dimensional (3D) scans of hemp stalks. These traits include stem length, chord length, diameter, node count, and node spacing, which are essential for characterizing plant architecture and supporting breeding, production, and processing research. In addition to leveraging existing 3D scan datasets and ground-truth measurements, this objective may include the design and development of an improved 3D imaging system that streamlines data acquisition and reduces reliance on proprietary software, thereby increasing accessibility and throughput.
Objective 2 is to expand and operationalize an existing digital framework for industrial hemp research that standardizes terminology, data structures, and data collection workflows. Building on prior research and ongoing system development, this effort will further integrate geospatial, phenotypic, and experimental data using existing enterprise technologies and cloud infrastructure. The objective seeks to enhance data consistency, interoperability, and accessibility across research programs while improving the efficiency of data collection, management, and analysis. Together, these objectives will support a more scalable and collaborative research ecosystem for industrial hemp.
Approach:
To achieve the first objective, the project will develop and validate a computational workflow for processing and analyzing three dimensional (3D) representations of hemp stems from structured light scanning systems. The workflow will use non proprietary file formats to ensure flexibility and long term independence from commercial platforms. Initial efforts will focus on preprocessing point cloud and mesh data, including noise reduction, alignment, normalization, and segmentation of individual stems. Subsequent analysis will apply geometric and computational methods such as skeletonization, centerline extraction, and surface interrogation to quantify structural features. Algorithms will detect nodes, calculate internode spacing, and derive curvilinear and chord lengths to assess stem curvature.
Automated measurements will be evaluated against manually collected ground truth data. Statistical analyses will assess accuracy, precision, and bias, with iterative refinements to improve performance. The workflow will support high throughput processing, enabling large scale analysis and compatibility with high performance and cloud computing environments.
In parallel, the project will explore improvements to upstream data acquisition by evaluating a streamlined 3D imaging system tailored for hemp phenotyping. This may include optimizing hardware configurations, imaging protocols, and data export processes to reduce inefficiencies. The goal is an integrated “scan to analysis” system enabling rapid, repeatable phenotyping at scale.
To address the second objective, the project will extend an existing digital data framework by refining standardized terminology, metadata schemas, and data collection protocols for industrial hemp research. Work will focus on harmonizing phenotypic, environmental, and experimental data across sources. Digital tools will be enhanced and deployed to capture structured, georeferenced data in field and laboratory settings.
Cloud infrastructure will support centralized storage, integration, and processing. Data pipelines will incorporate outputs from the automated 3D workflow alongside complementary datasets, emphasizing interoperability through standardized formats, application programming interfaces (APIs), and FAIR (Findable, Accessible, Interoperable, and Reusable) data principles. Advanced visualization and analytics tools will enable exploration of integrated datasets through interactive dashboards and geospatial applications. By expanding digital infrastructure and integrating automated phenotyping, the project will improve efficiency, consistency, and collaboration in industrial hemp research.