Crop Classification Dataset from Fused Optical and Radar Satellite Data

A dataset for crop mapping created by fusing optical and radar satellite imagery. The data originates from the UCI Machine Learning Repository, though specific details on volume, creator, and collection date are unavailable. It is designed for developing and testing classification models in agricultural remote sensing.

Use Cases

• Classifying crop types using fused spectral bands from optical sensors and backscatter values from radar.
• Comparing the performance of optical-only, radar-only, and fused data for land cover mapping tasks.
• Training models to predict agricultural land use from satellite-derived vegetation indices and radar texture features.
• Benchmarking image fusion techniques for improving the accuracy of crop classification algorithms.

Strengths

• Integrates data from two distinct satellite sensor modalities (optical and radar).
• Hosted on the UCI Machine Learning Repository, a known source for benchmark datasets.

Limitations

• The total number of samples, image resolution, and geographic coverage are unknown.
• Lacks metadata on specific crops, sensor types, and acquisition dates, limiting reproducibility.

Provenance

Source

UCI Machine Learning Repository

Collection Method

Fusion of optical and radar satellite imagery.