#approximation #distance #projection #flat #system #calculations #geospatial

flat_projection

Fast geodesic distance approximations via flat surface projection

5 releases (3 breaking)

Uses old Rust 2015

0.4.0 Sep 14, 2020
0.3.0 Sep 20, 2019
0.2.0 Apr 12, 2018
0.1.1 Jan 28, 2018
0.1.0 Jan 28, 2018

#116 in Geospatial

Download history 51/week @ 2023-12-16 107/week @ 2023-12-23 42/week @ 2023-12-30 25/week @ 2024-01-06 22/week @ 2024-01-13 4/week @ 2024-01-20 10/week @ 2024-01-27 30/week @ 2024-02-03 47/week @ 2024-02-10 26/week @ 2024-02-17 40/week @ 2024-02-24 50/week @ 2024-03-02 106/week @ 2024-03-09 66/week @ 2024-03-16

265 downloads per month

MIT license

18KB
131 lines

flat-projection

Build Status

Fast geodesic distance approximations via flat surface projection

The FlatProjection struct can by used to project geographical coordinates from WGS84 into a cartesian coordinate system. In the projected form approximated distance and bearing calculations can be performed much faster than on a sphere. The precision of these calculations is very precise for distances up to about 500 km.

Usage

extern crate flat_projection;

use flat_projection::FlatProjection;

fn main() {
    let (lon1, lat1) = (6.186389, 50.823194);
    let (lon2, lat2) = (6.953333, 51.301389);

    let proj = FlatProjection::new(51.05);

    let p1 = proj.project(lon1, lat1);
    let p2 = proj.project(lon2, lat2);

    let distance = p1.distance(&p2);
    // -> 75.648 km
}

Benchmark

$ cargo bench

distance/flat           time:   [322.21 ps 323.82 ps 326.41 ps]
distance/haversine      time:   [12.604 ns 12.831 ns 13.162 ns]
distance/vincenty       time:   [346.79 ns 348.00 ns 349.61 ns]

According to these results the flat surface approximation is about 40x faster than the Haversine formula.

License

This project is released under the MIT license.

Dependencies

~155KB