PhenoBench: UAV Images of Sugar Beet Crops and Weeds

PhenoBench is a large UAV image dataset for semantic image interpretation in agriculture, focusing on sugar beet crops and weeds. The dataset was introduced by Weyler et al. in a 2024 IEEE TPAMI paper. This release packages a subset of 2,179 samples from the original PhenoBench as a FiftyOne dataset.

Use Cases

• Train semantic segmentation models to distinguish between crops and weeds based on UAV imagery.
• Benchmark agricultural image interpretation algorithms on a standardized dataset.
• Develop weed detection systems for precision herbicide application.
• Analyze crop health and field conditions from aerial imagery.

Strengths

• Dataset contains 2,179 samples.
• Based on a large, peer-reviewed dataset introduced in a 2024 IEEE TPAMI publication.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the full dataset is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Voxel51 (packager). Original data from PhenoBench by Weyler et al.

Collection Method

UAV imagery collection over agricultural fields.

Freshness

Last updated 2026-05-21 21:47:56; freshness should be verified.