search

hamidrezaomidvar / LINDER

LINDER (Land use INDexER) is an open-source machine-learning based land use/land cover (LULC) classifier using Sentinel 2 satellite imagery
13 stars 2 forks source link

Adding roads from OSM #7

Closed hamidrezaomidvar closed 5 years ago

hamidrezaomidvar commented 5 years ago

I did a quick test, and it is possible to incorporate OSM road data to the current land cover. We need to specify the width of different roads based on their functionality. It is possible to calibrate it very easily for different cities. Here are example of OSM (top figure) and the road data I generated (black lines bottom figure)

Screenshot 2019-08-01 at 19 21 48 Screenshot 2019-08-01 at 19 21 44
sunt05 commented 5 years ago

This reminds me of sending you the chongqing data.

hamidrezaomidvar commented 5 years ago

Here if we combine roads and building data from OSM as well as GUF and prediction data for a part of London. Note that if all building data are available, then there would be another category which is not part of roads or buildings but still is paved (purple in this map), and basically can be merge to road and paved class. Note that in OSM, the map of roads are almost complete, therefore, for the places that the building footprint is not available or not complete, we can only use road footprint and assume the rest of built class to be buildings.

Screenshot 2019-09-04 at 11 47 51