Python utility to generate GeoJSON from photos stored in Apple's HEIC format. Turns an iPhone into an on the fly georecording device for quickly mapping locations during on-the-beat reporting.

Possible applications

As a student at Columbia's Lede Program, I quickly generated a map of all the trees in a three block radius in Manhattan by walking the streets and snapping a photo next to each trunk. I then dropped the photos in a folder, ran the script, and uploaded the GeoJSON to datawrapper.

I have also used a similar technique to map real-time reporting from student protests and plot warning bollards placed to mark an underground gas pipeline near campus.

Quickstart

Navigate to your project directory and clone this repository by running:

git clone https://github.com/declanrjb/photo-locator

In a new Python script or the interpreter, run:

import photolocator as pl

By default, locate_photos will return a complete GeoJSON as raw text. Each photo is represented as a point geometry with non-coordinate properties from the original file stored under properties.

pl.locate_photos('<photo_dir>')

The optional outpath flag writes the GeoJSON to a specified file path.

pl.locate_photos('<photo_dir>', '<out_name>.geojson')

GeoJSON can then be reformatted for the user's preferred mapmaking tool using geojson.io.

Worked Examples

Using arbitrary photos stored in demo/samples:

import photolocator as pl
pl.locate_photos('demo/samples', outpath='demo/photos.geojson')

The visualized data shows that the photos were succesfully located: