Remote Sensing Scenes Segmentation Dataset
The "Remote Sensing Scenes Segmentation Dataset" is a specialized collection for the remote sensing domain, comprised of high-resolution satellite images sourced from the internet, with dimensions ranging from 10752 x 10240 to 12470 x 13650 pixels. This dataset is designed for semantic segmentation, with annotations covering various natural and man-made features such as buildings, forests, water bodies, roads, and farmland.
Sample


Specification
Dataset ID
MD-Image-030
Dataset Name
Remote Sensing Scenes Segmentation Dataset
Data Type

Volume
About 100
Data Collection
Internet collected remote sensing satellite images, with resolution range from 10752 x 10240 to 12470*13650 pixels.
Annotation
Semantic Segmentation
Annotation Notes
The objects include buildings, forests, water bodies, roads, farmland, background.
Application Scenarios
Remote Sensing
Data Collections
This compact dataset consists of 100 satellite images, each meticulously annotated to identify key elements like buildings, forests, water bodies, roads, and agricultural areas. The high resolution of the images ensures detailed and accurate segmentation, making this dataset an essential tool for applications in environmental monitoring, urban planning, and land use analysis, providing insights into the composition and changes of earth's landscapes.
Quality Assurance

Related products
Any further information, please contact us.