UAV Phenotyping Data for Megathyrsus maximus with Machine Learning Analysis

A study by Guilherme Francio Niederauer, published on figshare in 2026, provides tabular data from UAV-based phenotyping of a Megathyrsus maximus (guinea grass) biparental population. The dataset likely contains results from optimizing image acquisition parameters, evaluating digital traits like pixel count and Haralick's entropy, and applying machine learning models to predict yield and canopy height. It includes analyses of ground sampling distance effects, heritability estimates, and genotype selection efficiency.

Use Cases

• Optimizing UAV image acquisition parameters for forage crop phenotyping based on ground sampling distance analysis.
• Predicting green and dry matter yield in forage breeding programs using machine learning models on RGB-derived digital traits.
• Evaluating the heritability and environmental susceptibility of digital traits like pixel count and vegetative indices.
• Comparing genotype selection efficiency between UAV-derived digital traits and conventional yield measurements.

Strengths

• Includes machine learning model results with strong predictive power (r > 0.80) for yield traits.
• Reports high broad-sense heritability values (0.7 to 0.87) for key yield-related digital traits.
• Provides analysis of optimal ground sampling distance resolutions (0.5–1.0 cm) for UAV phenotyping.
• Demonstrates 80% correspondence between top genotypes ranked by UAV-derived pixel count and conventional yield measurements.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is small in scale (427.3 KB), indicating limited raw observational data.

Provenance

Source

figshare

Collection Method

Data likely originates from a controlled UAV-based phenotyping experiment on a Megathyrsus maximus biparental population, with subsequent machine learning and mixed model analysis.

Time Range

The study was published in 2026; the experimental timeframe is not specified.

Freshness

Last updated 2026-04-22 05:46:06; freshness should be verified.

Geography

The geographic scope of the experiment is not specified in the provided metadata.