Convert PDF to KML

A map stored as a PDF must be converted when it needs to open in Google Earth, be shared as location-aware data, or be edited as geographic features. The result is meaningful only if the PDF contains spatial reference information or can be accurately georeferenced.

What the PDF format is

PDF is a fixed-layout document format used for maps, survey plans, reports, permits, and GIS or CAD printouts. It can contain text, raster images, vector drawing commands, layers, and— in a GeoPDF—geospatial metadata. A PDF’s visual appearance does not prove that it contains geographic data.

A scanned or exported map is often only a page image with no coordinate system. Vector lines in a PDF may be ordinary drawing objects with no feature attributes, topology, or usable coordinates. A GeoPDF may contain a coordinate reference system and map extent that GIS software can read.

What the KML format is

KML is an XML-based geographic format supported by Google Earth, Google Earth Pro, Cesium, and many GIS applications. It stores points, lines, polygons, labels, descriptions, styles, image overlays, and optional altitude values for 3D display.

KML normally uses geographic coordinates in WGS 84, EPSG:4326. .kmz is a ZIP-based KML package that can include icons, images, and other referenced files. KML is suited to interactive viewing and sharing, but it does not preserve every PDF layout feature or GIS project setting.

How to convert the file

QGIS for georeferencing and digitizing

QGIS is the most practical free desktop option because it can open many geospatial PDFs, georeference ordinary map images, digitize features, and export vector layers to KML.

1. Install a current QGIS release and work from a copy of the PDF.
2. Try Layer → Add Layer → Add Raster Layer and select the PDF. If QGIS reads its geospatial reference and places it correctly, record the detected CRS. If the PDF has multiple pages or does not load, render the required page to PNG or TIFF first.
3. For an unreferenced scan or page image, open Raster → Georeferencer. Add the image, choose a target CRS, and create control points at identifiable locations such as coordinate-grid intersections or surveyed road junctions. An affine transformation needs at least three non-collinear points; use more points distributed across the map and inspect the residual error. Save the result as a georeferenced GeoTIFF.
4. Create a temporary point, line, or polygon layer with Layer → Create Layer → New Shapefile Layer or a GeoPackage layer. Set its CRS to the georeferenced map’s CRS and add fields for names or descriptions.
5. Enable editing and trace the required roads, boundaries, points, or areas. A PDF’s visible lines are not automatically converted into GIS features, so this digitizing step is usually manual.
6. Right-click the completed vector layer and choose Export → Save Features As. Set Format to Keyhole Markup Language (KML), set the output CRS to EPSG:4326, and save a file such as converted.kml.
7. Open the KML in Google Earth Pro and check alignment, geometry, labels, and styles. To package referenced images and icons, use File → Save Place As and select KMZ where appropriate.

If QGIS exposes usable vector layers from a GeoPDF, export those layers instead of tracing them. Many GeoPDFs still require layer selection, cleanup, or manual digitization because PDF graphics do not necessarily retain GIS attributes or topology.

GDAL after extracting GIS data

GDAL can convert an already extracted GIS layer to KML, but it is not a reliable PDF feature-recovery tool. For example, after converting the content to GeoPackage, inspect its layers and reproject them during export:

ogrinfo extracted.gpkg

ogr2ogr -f KML converted.kml extracted.gpkg -t_srs EPSG:4326

Specify a source layer when the GeoPackage contains more than one layer, and verify that the source CRS is correctly defined before reprojecting.

Online converters

MyGeodata Cloud can be tested with a GeoPDF or another PDF that contains machine-readable spatial data. Upload the file, select the detected layer and coordinate system, export KML, and inspect the result in Google Earth. Online conversion is unsuitable for confidential plans, permits, or survey data unless the service’s data-retention terms are acceptable.

Generic document converters often rasterize the PDF or return an empty or poorly located KML. No converter can infer accurate geographic coordinates from an ordinary page image without known control points. If the service cannot show the detected CRS, layer content, or output extent, use QGIS instead.

Quality and compatibility limits

• A scanned or print-only PDF needs georeferencing and usually manual digitization. OCR can recover text labels but cannot establish feature geometry or coordinates.
• PDF conversion commonly loses GIS attributes, feature IDs, topology, layer hierarchy, fonts, annotations, and complex symbology. Recreate important names and classifications in the GIS layer.
• Use control points across the entire map, not in one corner. Compare exported features with known roads, boundaries, or coordinates and review the transformation error.
• A wrong source or target CRS can shift every feature by a large distance. KML export normally requires WGS 84 geographic coordinates, EPSG:4326.
• Google Earth may render KML styles, labels, transparency, and altitude differently from the source PDF or GIS project. Test the file in the intended viewer.
• For a large map that only needs to be displayed, a georeferenced raster image overlay packaged in KMZ may be more appropriate than tracing thousands of KML vertices. It remains an image overlay, not editable geographic features.