Dataset Hub

The GLH-Bridge dataset is a large‑scale dataset for bridge detection in large‑size very high resolution (VHR) remote sensing images. More details can be found in the paper "Learning to Holistically Detect Bridges from Large‑Size VHR Remote Sensing Imagery" (TPAMI 2024).

PatternNet is a large‑scale high‑resolution remote sensing dataset for scene classification and image retrieval. It contains 38 classes, each with 800 images of size 256×256 pixels, totaling 30,400 images. The images are sourced from Google Earth or Google Map API, covering various land‑cover types such as airplanes, baseball fields, basketball courts, beaches, bridges, etc. The image resolution is 256 × 256 m with three RGB bands.

This dataset includes aerial imagery and Sentinel‑2 imagery. The aerial imagery has dimensions 512×512×5, spatial resolution 0.2 m, and contains 5 channels (RGB, NIR, elevation). Sentinel‑2 imagery has spatial resolution of 10–20 m, includes 10 spectral bands, snow/cloud mask probability range 0–100, and temporal step format T×10×W×H. The label (mask) size is 512×512, includes 13 classes covering various types from buildings to other unclassified land covers.

The dataset contains five open‑access polarimetric SAR images from the San Francisco area, acquired by different satellites at different times. The images have high scientific value, are pixel‑level annotated for classification and segmentation, and are released under an open‑source license.

The NWPU VHR‑10 dataset is a challenging benchmark for geospatial object detection containing ten categories. It includes 800 very‑high‑resolution (VHR) optical remote‑sensing images: 715 color images from Google Earth (spatial resolution 0.5–2 m) and 85 panchromatic‑sharpened color‑infrared images from Vaihingen (0.08 m). The dataset is split into a positive set (650 images containing at least one target) and a negative set (150 images with no targets). The positive set is manually annotated with 757 aircraft, 302 ships, 655 storage tanks, 390 baseball fields, 524 tennis courts, 159 basketball courts, 163 athletics fields, 224 ports, 124 bridges, and 477 vehicles using bounding boxes and instance masks as ground truth.

HRSC2016: Ship detection dataset in remote sensing images. DIOR: Object detection dataset in remote sensing images. HRSID: Small object detection dataset in remote sensing images. NWPU VHR‑10 dataset: Object detection dataset in remote sensing images.

FAIR1M is a fine‑grained object recognition and detection dataset focusing on high‑resolution (0.3‑0.8 m) RGB images sourced from Gaofen satellites and Google Earth. It contains 15 000 high‑resolution images covering five major categories (ships, vehicles, aircraft, ball‑fields, roads) and 37 sub‑categories. Annotations are provided as rotated bounding boxes, suitable for remote sensing, Earth observation, geospatial, and satellite‑image research.

This is an open remote‑sensing image object detection dataset containing objects such as aircraft, oil tanks, playgrounds, and overpasses. The dataset follows the PASCAL VOC format and consists of four files, each corresponding to one object type.