PASTIS-HD: Panoptic Agricultural Satellite Time Series

French metropolitan territory agricultural parcels annotated for panoptic segmentation. The dataset contains 2,433 patches, each representing a Sentinel-2 multispectral image time series of variable length. It was authored by IGNF and last updated on the Hugging Face platform in August 2025.

Use Cases

• Train panoptic segmentation models based on pixel-level instance and semantic labels.
• Develop semantic segmentation algorithms for agricultural land use from satellite imagery.
• Analyze crop patterns and parcel boundaries using very high-resolution satellite data.
• Benchmark time-series analysis models for multispectral satellite imagery.
• Research agricultural monitoring techniques combining optical and radar time series data.

Strengths

• Contains 2,433 annotated patches, providing a substantial number of samples.
• Provides panoptic annotations, which include both an instance index and a semantic label for each pixel.
• Includes multiple data modalities: optical time series, radar time series, and very high-resolution images as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Data may reflect geographic bias inherent to its collection within the French metropolitan territory.

Provenance

Source

IGNF

Freshness

Last updated 2025-08-19 07:35:02; freshness should be verified.

Geography

French metropolitan territory