Convex Hull — GeoJSON, Shapefile, KML and more
A convex hull is the smallest convex polygon that encloses a set of geographic features. Use these free browser-based tools to compute the convex hull of any vector file — points, lines, or polygons — in any common format, without installing software.
What is a convex hull?
A convex hull (also called a convex envelope or convex closure) is the smallest convex polygon that contains all features in a dataset. Imagine stretching a rubber band around all your features and letting it snap — the shape the rubber band forms is the convex hull.
The algorithm works by finding the outermost coordinates across all features and computing the minimal bounding polygon. Unlike a bounding box, which always produces a rectangle, the convex hull fits the actual shape of your data. The result is always a single Polygon feature regardless of how many input features were provided.
Convex hulls are computed entirely in the browser using Turf.js. No data is ever sent to a server.
Common use cases
Study area delineation
Automatically define the geographic extent of a field survey, species sighting dataset, or point collection as a polygon.
Clustering visualization
Visually group clusters of points by computing the convex hull of each cluster — useful for showing the extent of customer zones, animal territories, or event areas.
Spatial index pre-filtering
Use the convex hull as a rough bounding geometry to pre-filter candidates in a spatial query before running a more expensive precise intersection.
Coverage checking
Compare the convex hull of a dataset against a reference area to identify gaps or ensure complete coverage.
How to use
- 1
Upload your vector file (GeoJSON, Shapefile ZIP, KML, GeoPackage, or GML).
- 2
The file is parsed in your browser using GDAL WebAssembly and converted to GeoJSON.
- 3
Turf.js computes the convex hull — the smallest convex polygon enclosing all features.
- 4
The result is converted back to your chosen format and downloaded.
Frequently asked questions
What geometry types does this work on?
Any geometry type — points, lines, and polygons are all supported. The convex hull is computed across all coordinates in the file regardless of geometry type.
Why does the tool say it cannot compute a convex hull?
A convex hull requires at least 3 non-collinear points. If your file has fewer than 3 features, or all features lie on a straight line, the convex hull cannot be computed.
The convex hull always produces a single polygon — is that correct?
Yes. The convex hull is by definition a single polygon that encloses all input features. If you want a hull per cluster or per attribute group, you would need to split your data first.
How is this different from a bounding box?
A bounding box is always an axis-aligned rectangle. A convex hull fits the actual shape of your data and will generally be a tighter, more useful boundary.