Meta_Album_RSICB_Mini: Remote Sensing Land Use Images

A collection of 1,800 satellite and aerial images for remote sensing classification, resized to 128x128 pixels. The dataset covers 45 scene categories organized into a 2-level hierarchy with 6 super-categories, including agricultural land and construction land. It was created by Haifeng Li et al. and preprocessed for the Meta-Album benchmark in March 2022.

Use Cases

• Training few-shot image classifiers based on the 45 land use scene categories.
• Benchmarking meta-learning algorithms for remote sensing based on the standardized image format and class hierarchy.
• Developing land cover classification models based on satellite and aerial imagery.
• Studying hierarchical classification methods based on the 6 super-categories of land use.

Strengths

• Contains 1,800 images across 45 distinct classes, providing a multi-class benchmark.
• Images are preprocessed to a consistent 128x128 pixel resolution using an anti-aliasing filter.
• Features a 2-level label hierarchy with 6 super-categories, adding semantic structure.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count is relatively small (1,800 images), which may limit suitability for large-scale training.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

RSI-CB: A Large Scale Remote Sensing Image Classification Benchmark via Crowdsource Data (GitHub: lehaifeng/RSI-CB).

Collection Method

Images were assembled from various global locations and preprocessed for the Meta-Album benchmark.

Geography

Various locations around the world; classification follows China's landuse standard.