The team chose UgCS and its Circlegrammetry tool to fly automated orbiting missions around individual vine sections with a DJI Mavic 3 Multispectral drone.
The Research Challenge: Vineyard Canopy Occlusion
Vineyard canopies hide what's happening at the base of the vines, creating data occlusion. Traditional top-down photogrammetry surveys capture canopy health but miss the lower portions of the plant where critical growth patterns develop.
To build accurate 3D models with multispectral data at all levels of the vine, researchers needed:
- Automated circular flight paths around target areas
- Consistent camera angles and overlap for 3D reconstruction
- Multispectral data capture throughout the orbit
- Repeatable missions for tracking changes over time
- Precise altitude and speed control for uniform data quality
Manual flight wouldn't work. The orbits needed exact positioning, consistent speed, and coordinated camera control to generate the point cloud density required for Gaussian splatting reconstruction.
The Solution: UgCS Circlegrammetry for Precision Agricultural Research
UgCS Circlegrammetry uses automated circular flight patterns to effectively perform an oblique scan of a selected map region area. The team used it to orbit vine sections at multiple altitudes and angles, capturing images from all sides.
Key features used:
- Circlegrammetry tool for automated 360-degree orbits with adjustable radius and altitude
- DJI Mavic 3 Multispectral support for simultaneous RGB and multispectral band capture
- Desktop mission planning for precise parameter control before heading to the field
- Repeatable flight paths to track vine development across growing seasons
- UgCS Open (free version) to test and validate the methodology
The circular missions capture vine structure from angles overhead mapping can't reach. Combined with multispectral sensors, this reveals how light penetration, irrigation, and other factors affect growth at different canopy levels.
Research Results: Managinging Drone Prop Wash in Low-Altitude Mapping
The Circlegrammetry missions delivered the 360-degree coverage the team needed, but low-altitude flights introduced an unexpected problem. Rotor downwash (prop wash) from flying close to the vines created motion blur in the captured images, preventing the high-quality 3D Gaussian splatting reconstruction that the research required.
The team is now refining their approach. Upcoming flights will test higher altitudes with adjusted camera angles to maintain detail while avoiding prop wash interference. One option under consideration is using a telephoto lens at an increased altitude to preserve ground sampling distance without the aerodynamic disturbance.
Winter phase: mapping dormant vine structure
During the dormant season when leaves drop, the team plans to survey exposed vine canes to identify structural health issues. With over 200,000 individual vine stocks in the vineyard, manual inspection isn't practical. Drone surveys can systematically identify vines with structural problems that need replacement before the next growing season.
Current outcomes:
- Validated automated circular flight paths for vineyard research
- Identified altitude constraints due to prop wash effects
- Established repeatable methodology for testing different flight parameters
- Demonstrated scalability across large vineyard areas
The initial results revealed real-world constraints that weren't obvious during planning. The team now has a clear path forward for achieving the data quality their research demands.
Technical Specifications
UAV Platform: DJI Mavic 3 Multispectral (Mavic 3M)
Flight planning software: UgCS Enterprise (using UgCS Open for initial testing)
Mission type: UgCS Circlegrammetry (automated circular orbits)
Data type: RGB + multispectral bands
3D Processing: 3D Gaussian splatting reconstruction & Point Cloud Generation
Institution: University of California San Diego
Application: Precision agriculture research
Methodology: Multi-altitude orbital flights to capture complex vine geometry for 3D reconstruction
Ongoing Research and Development
The UC San Diego team continues to develop this methodology across multiple growing seasons. The project remains active, with researchers expanding data collection to different vineyard blocks and refining their 3D reconstruction techniques.
Current work includes testing different orbit parameters for various vine types, comparing multispectral data from different canopy positions, and validating the Gaussian splatting approach against ground-truth measurements. The team is also exploring how temporal changes captured through repeated Circlegrammetry missions can predict harvest outcomes.
The research team is working closely with UgCS to optimize flight parameters for agricultural research applications. This collaboration helps refine both the data collection methodology and the software tools that support it. As the project progresses, findings will inform best practices for multi-angle agricultural drone surveys and contribute to precision viticulture research.
Future phases of the project will scale the approach to larger vineyard areas and investigate whether the same techniques apply to other crop types where canopy structure affects growth and yield.
This type of multi-angle multispectral capture opens new possibilities for understanding vine health, growth patterns, and response to management practices. The ongoing nature of the research means the methodology will continue to evolve based on real-world field results and processing refinements.
Try UgCS Circlegrammetry for your agricultural research or commercial vineyard monitoring projects.