language:
- en
license: apache-2.0
size_categories:
- 1M<n<10M
task_categories:
- image-to-text
- text-to-image
- visual-question-answering
- zero-shot-classification
- summarization
pretty_name: RSTeller
tags:
- remote-sensing
- vision-language model
- LLM
Metadata for RSTeller
This dataset contains the necessary metadata for the dataset SlytherinGe/RSTeller.
Dataset Details
Description
The metadata table provides detailed information for the RSTeller dataset, with the following columns:
patch_id: The primary key of the table, corresponding to the"__key__"or the"patch_id"field in the JSON of the RSTeller dataset.patch_latandpatch_lon: The latitude and longitude coordinates of the patch center in WGS84 format.osm_id: The OpenStreetMap (OSM) ID of the element used for captioning. This can be used as an identifier to query raw information from OSM via Overpass.osm_type: Indicates whether the OSM element is classified as "area" or "non-area."osm_tags: The raw, unfiltered tags associated with the OSM element. These tags are serialized in JSON format. Note that the tags reflect the state of OSM one month after the image was captured and may not be up-to-date in Overpass.osm_geometry: The Well-Known Text (WKT) representation of the OSM element's geometry in the WGS84 coordinate system. The geometry is cropped to exclude regions outside the image patch.osm_geometry_norm: The normalized geometry of the OSM element, where(0, 0)represents the bottom-left corner and(1, 1)represents the top-right corner of the image patch.image_name: The name of the original image from which this patch was cropped. This field can be used to query data from Google Earth Engine.
This metadata serves as a critical resource for aligning OSM elements with their corresponding remote sensing image patches in RSTeller.