Use GraphHopper's map-matching lib from matlab (a java wrapper)
first, compile this project and import map file into a dir. (You have to done this by Java directly because doing this in matlab may cause some err).
java -cp "*" edu.buaa.ACTMapMatching D:\map-matching-matlab\beijing_around.pbf D:\map-matching-matlab\cachethen, modify matlab classpath file to allow seek of the map-matching lib. Type edit classpath.txt in matlab console,
paste following to the file:
# D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/commons-compress-1.10.jar
# D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/commons-io-1.3.1.jar
# D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/commons-logging-1.0.4.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/graphhopper-core-0.8.2.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/graphhopper-map-matching-core-0.8.2.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/graphhopper-reader-osm-0.8.2.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/graphhopper-tools-lgpl-0.8.2.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/hmm-lib-1.0.0.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/map-matching4matlab.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/osmosis-osm-binary-0.44.1.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/protobuf-java-2.6.1.jar
# D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/slf4j-api-1.7.7.jar
D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/trove4j-3.0.3.jar
# D:/map-matching-matlab/out/artifacts/map_matching4matlab_jar/xmlgraphics-commons-2.1.jar#
# D:\matlabcalljava\out\artifacts\matlabcalljava_jar\matlabcalljava.jarsome are commented because matlab already have include such jar (although it could be a different version). restart matlab to make above config work.
finally, call java methods from matlab. (check example.m for example)
t = edu.buaa.ACTMapMatching
t.load('D:\map-matching-matlab\cache', 20); % 20 is the gps accuracy, unit one meter
% example trajectory, first column is latitude, second is longitude, third is timestamp (seconds).
trajectory=[
39.862180 116.471353 1478793726
39.862155 116.471312 1478793732
39.862398 116.471278 1478793738
39.862892 116.471308 1478793744
39.862892 116.471308 1478793750
39.863462 116.471302 1478793756
39.864285 116.47125 1478793762
39.864728 116.471197 1478793768
39.864728 116.471197 1478793774
39.865840 116.47124 1478793780
39.865840 116.47124 1478793786
39.866227 116.471272 1478793792
39.866227 116.471272 1478793798
39.866227 116.471272 1478793804
39.866227 116.471272 1478793810
39.866227 116.471272 1478793816
39.866227 116.471272 1478793822
39.866227 116.471272 1478793828
39.866227 116.471272 1478793834
39.866227 116.471272 1478793840
39.866683 116.471233 1478793846
39.867257 116.471232 1478793852
39.867898 116.471197 1478793858
39.868600 116.471212 1478793864
39.869195 116.47123 1478793870
39.869748 116.471265 1478793876
39.869902 116.471512 1478793882];
road_path = t.toRoads(trajectory);
% you got something like this,
%
% 0 1478793829.489 8733 73 856
% 0 1478793878.839 8734 49 256
%
% the second column is the timestamp (second) at which the car enter the road,
% the first column is the timeslot the timestamp falls (48 timeslot a day)
% the third column is roadId which is only unique for one import, if you import map a second time, it may change.
% the fourth column is travelTime which is the time the car spent to travel through the road (seconds).
% the fifth column is road length (meters).
% if you want raw output from map-matching lab, try:
roads = t.rawMM2Roads(trajectory);
% you got a list of roads, each roads contains serval gps points
% [
% [39.862180 116.471353],
% [39.862180 116.471353],
% [39.862180 116.471353]
% ],[
% [39.862180 116.471353],
% [39.862180 116.471353],
% [39.862180 116.471353]
% ],
% ...
%
matched = t.exactPoints(trajectory);
% you got something like this, not all points are matched.
% road_id, matched_lat, matched_lon, origin_lat, origin_lon, origin_time
% 91558 39.8621799403485 116.471331237138 39.8621800000000 116.471353000000 1478793726.00000
% 91558 39.8628920583424 116.471329285239 39.8628920000000 116.471308000000 1478793744.00000
% 91558 39.8634620705058 116.471327722848 39.8634620000000 116.471302000000 1478793756.00000
% 8733 39.8642863311037 116.471309927235 39.8642850000000 116.471250000000 1478793762.00000
% 8733 39.8647302893028 116.471300066036 39.8647280000000 116.471197000000 1478793768.00000
% 8733 39.8658407862976 116.471275399677 39.8658400000000 116.471240000000 1478793780.00000
% 8733 39.8662268850230 116.471266823652 39.8662270000000 116.471272000000 1478793792.00000
% 8734 39.8666830289700 116.471226998921 39.8666830000000 116.471233000000 1478793846.00000
% 8734 39.8672570107662 116.471229769803 39.8672570000000 116.471232000000 1478793852.00000
% 8734 39.8678978268708 116.471232863325 39.8678980000000 116.471197000000 1478793858.00000
% 8734 39.8685998829218 116.471236252482 39.8686000000000 116.471212000000 1478793864.00000
% 8734 39.8691949559484 116.471239125180 39.8691950000000 116.471230000000 1478793870.00000
% 8734 39.8697481120190 116.471241795526 39.8697480000000 116.471265000000 1478793876.00000
% 57875 39.8698396722739 116.471486794774 39.8699020000000 116.471512000000 1478793882.00000
matched = t.simulatedTraj(trajectory);
% you got something like above, not all points are matched.
% also matched_lat, matched_lon may have more points (origin_lat, origin_lon is -1, origin_time is linear random inserted),
% to form a complete trajectory for route.
% Export trajectory to a gpx file (for visualization in Google Earth).
t.toGPXFile(trajectory, "file-path-here.gpx");
% K-d anonymous
% input data columns: car_id, time_aligned, latitude, longitude
% output is the same format.
% Example:
%
trajectories=[
1 1 39.862180 116.47135
1 2 39.862155 116.471312
...
2 1 39.990 116.470
2 2 39.999 116.488
...
3 1 39.155 116.155
3 2 39.156 116.178
...
];
anonymous_level_k = 2;
anonymous_trajectories = t.kdAnonymous(trajectories, anonymous_level_k)
[
1 1 39.86180 116.4735
1 2 39.86155 116.47312
...
2 1 39.9590 116.4170
2 2 39.9399 116.4588
...
3 1 39.1155 116.1455
3 2 39.1156 116.1278
...
]