airtiler

The airtiler generates training / test data for neural networks by downloading buildings from vector data from OpenStreetMap and the corresponding satellite images from Microsoft Bing Maps.

It then generates binary masks from the vector data which can be used for example for instance segmentation.

Examples

Instance Separation	Image	Mask
False
True

Installation

To install airtiler run:

pip install airtiler

Usage

airtiler -c sample_config.json

API

airtiler = Airtiler("bing_key")
airtiler.process(config)

Config

Key	Required
options
boundingboxes	Yes

Options (optional)

Key	Description
target_dir	The directory where the files will be written to
zoom_levels	Global zoom levels which will be used, if a boundingbox if specified in short format or has no boundingboxes.
separate_instances	If true, each building instance will be separated. Otherwise, a building consisting from multiple instances will be rendered as one.

Sample config

{
  "options": {
    "target_dir": "./output/blabla",
    "zoom_levels": [15, 16, 17],
    "separate_instances": false
  },
  "query": {
    "tags": ["highway", "building", "leisure=swimming_pool"]
  },
  "boundingboxes": {
    "firenze": [11.239844, 43.765851, 11.289969, 43.790065],
    "rapperswil": {
      "zoom_levels": [17, 18],
      "tr": 8.818724,
      "tl": 47.222126,
      "br": 8.847435,
      "bl": 47.234629
    },
    "new_york": {
      "tr": -74.02059,
      "tl": 40.646089,
      "br": -73.864722,
      "bl": 40.77413
    }
  }
}

Projects

The airtiler is used in the following projects:

Deep OSM - Instance segmentation using orthophotos and OSM data

Metadata

Release files for airtiler 2.1.7

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for airtiler 2.1.7
File	Size	Uploaded
airtiler-2.1.7.tar.gz	7.2 kB	May 17, 2018	Details

Release files / airtiler-2.1.7.tar.gz

Download URL	airtiler-2.1.7.tar.gz
Size	7.2 kB
Tags	Source
SHA-256 checksum How to use checksums	`e0803e5cd18461d6242f62febeb53153a92e4e991957e78509f1541635df841f`
BLAKE2b-256 checksum How to use checksums	`c82ed5764065d19b631abeec74dcfe574d22c7f0c5849f459881e96bd1eca599`
Upload date	May 17, 2018
Uploaded using Trusted Publishing? What is trusted publishing?	No