Bayesian Map-matching
Project description
bmm: Bayesian Map-Matching
Map-matching using particle smoothing methods.
Docs and methodology.
Install
pip install bmm
Load graph and convert to UTM
UTM (Universal Transverse Mercator) is a commonly used projection of spherical longitude-latitude coordinates into square x-y coordinates.
import numpy as np
import pandas as pd
import osmnx as ox
import json
import bmm
graph = ox.graph_from_place('Porto, Portugal')
graph = ox.project_graph(graph)
Load polyline and convert to UTM
data_path = 'simulations/porto/test_route.csv'
polyline_longlat = json.loads(pd.read_csv(data_path)['POLYLINE'][0])
polyline_utm = bmm.long_lat_to_utm(polyline_longlat, graph)
or generate fake data
fake_route, fake_polyline_utm = bmm.sample_route(graph, timestamps=15, num_obs=25)
Offline map-matching
matched_particles = bmm.offline_map_match(graph, polyline=polyline_utm, n_samps=100, timestamps=15)
Online map-matching
# Initiate with first observation
matched_particles = bmm.initiate_particles(graph, first_observation=polyline_utm[0], n_samps=100)
# Update when new observation comes in
matched_particles = bmm.update_particles(graph, matched_particles, new_observation=polyline_utm[1], time_interval=15)
Plot
bmm.plot(graph, particles=matched_particles, polyline=polyline_utm)
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
bmm-1.3.tar.gz
(40.9 kB
view hashes)