I'm having trouble understanding the concept of a Vector tile. My understanding of Vector is point, line, polygon. So buildings, roads, areas etc.

My understanding of Raster is a grid of pixels each with a value to describe it. So the actual imagery would be raster.

I don't understand how a client can request a vector tile and the "imagery" (which I'm used to being raster - .tiff,.png,.jpeg etc) be presented as a vector.

How are these raster images being "vectorized"?

How can points, lines and polygons create something as complex as a raster image?

1 Answer 1


A vector tile is still vector data as you understand it. It is just being delivered differently, by breaking it into chunks that fit into a tiling scheme. This speeds up the delivery because you are only getting data for the area in your view extent. Once the data is delivered, it gets rendered by the (browser) client.

  • So where does the imagery/basemap come from ? Which we usually layer vector data on top of Dec 11, 2018 at 19:21
  • 2
    The basemap or aerial imagery comes through as a (tiled) raster from a separate request, and the client combines them. Dec 11, 2018 at 20:56
  • Ohh ok, so "Vector Tiles" as a technology still uses Raster data? Why does everyone talk about the small size of vector tiles? Wouldn't they still be large if they are using rasters? Dec 11, 2018 at 21:03
  • 2
    Not quite. Vector tiles as a technology are separate and different from raster tiles. Web-based maps that use vector tiles for the overlay data will also use raster tiles for the base if you want the kind of visual effect that you get from images. You don't eliminate the size of the raster, but vector tiles are less data than non-tiled vector features. Dec 11, 2018 at 21:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.