Indoor map generation method and apparatus

1 . A computer-implemented method of indoor map generation, the method comprising:

identifying a plurality of first tracks based on information received from sensors of a plurality of terminal devices within a period of time and a pedestrian dead reckoning (PDR) algorithm, wherein each first track of the plurality of first tracks represents a curve comprising a position movement of a terminal device, of the plurality of terminal devices, within the period of time, and the first track comprises a plurality of step points of the terminal device on a same floor of a building in a process of the position movement;

for each first track of the plurality of first tracks,

segmenting the first track, to identify a set of fourth tracks, based on a change in information from at least one of the sensors, and

segmenting the first track, to identify a set of fifth tracks, based on one or more inflection points on the first track;

selecting, from sets of fourth tracks and sets of fifth tracks identified from the plurality of first tracks, a plurality of first target tracks, wherein an accuracy rate of the plurality of first target tracks is higher than an accuracy rate of a plurality of second tracks from the plurality of first tracks;

for each first target track, connecting the first target track to at least one second track corresponding to the first target track, to generate a branch, wherein the branch comprises the first target track and the at least one second track corresponding to the first target track; and

generating an indoor map for indoor positioning by combining the generated branches.

2 . The method according to claim 1 , wherein each first target track comprises a radio frequency fingerprint point that indicates a radio signal source obtained by one of the plurality of terminal devices through scanning in the process of the position movement, and a radio frequency fingerprint point comprised on the at least one second track corresponding to the first target track that matches a radio frequency fingerprint point on the first target track.

3 . The method according to claim 2 , wherein

the at least one second track comprises a radio frequency fingerprint point matching a first radio frequency fingerprint point comprised on the first target track.

4 . The method according to claim 1 , further comprising:

for each branch of the generated branches, correcting a coordinate system of the at least one second track to a coordinate system of the first target track corresponding to the at least one second track.

5 . The method according to claim 1 , wherein

each first track comprises a radio frequency fingerprint point that indicates a radio signal source obtained by one of the plurality of terminal devices through scanning in the process of the position movement; and

combining the generated branches comprises:

combining the generated branches based on mutually matched radio frequency fingerprint points in the generated branches.

6 . The method according to claim 1 , wherein there is a mapping relationship between an identifier of a radio frequency fingerprint point of the branch, signal strength, and a media access control (MAC) value, and the radio frequency fingerprint point indicates a radio signal source obtained by the terminal device through scanning in the process of the position movement.

7 . The method according to claim 6 , wherein mutually matched radio frequency fingerprint points of at least two branches of the generated branches correspond to a same MAC value, and a difference between signal strength of the mutually matched radio frequency fingerprint points falls within a preset range.

8 . The method according to claim 1 , wherein identifying the plurality of first tracks comprises:

identifying a plurality of third tracks based on the information received from the sensors in the plurality of terminal devices within the period of time and the PDR algorithm, wherein a third track represents a curve comprising the position movement of the terminal device in the period of time, and the third track comprises step points of the terminal device on a plurality of floors in the process of the position movement; and

segmenting each third track by floor, to obtain the plurality of first tracks.

9 . The method according to claim 1 , wherein the plurality of first target tracks are tracks whose scores are higher than a first threshold in the sets of fourth tracks, and a fourth track is part of a first track and whose accuracy rate is higher than a second threshold.

10 . The method according to claim 1 , wherein

the plurality of first target tracks are tracks whose scores are higher than a third threshold in the sets of fifth tracks; and

prior to segmenting the first track, to identify the set of fifth tracks, based on the one or more inflection points on the first track, the method further comprises:

detecting the one or more inflection points on the first track using a sliding window.

11 . The method according to claim 1 , wherein combining the generated branches comprises:

separately clustering a plurality of inflection points in a first map obtained by combining the generated branches, to obtain a clustering center of the plurality of inflection points, wherein the plurality of inflection points are located at a connection point of the generated branches in the first map;

for a first inflection point, traversing a track passing through the first inflection point, to identify a second inflection point adjacent to the first inflection point, wherein the first inflection point is an inflection point in the first map, and a clustering center of the first inflection point is a first clustering center; and

correcting a track between the first inflection point and the second inflection point based on the first clustering center and a second clustering center, wherein the second clustering center is a clustering center of the second inflection point.

12 . The method according to claim 1 , wherein the information collected by the sensors in the plurality of terminal devices within the period of time comes from crowdsourcing data collected by the plurality of terminal devices.

13 . An indoor map generation device, comprising:

at least one processor; and

at least one memory coupled to the at least one processor to store program instructions, which when executed by the at least one processor, cause the at least one processor to:

identify a plurality of first tracks based on information received from sensors of a plurality of terminal devices within a period of time and a pedestrian dead reckoning (PDR) algorithm, wherein each first track of the plurality of first tracks represents a curve comprising a position movement of a terminal device, of the plurality of terminal devices, within the period of time, and the first track comprises a plurality of step points of the terminal device on a same floor of a building in a process of the position movement;

for each first track of the plurality of first tracks,

segment the first track, to identify a set of fourth tracks, based on a change in information from at least one of the sensors, and

segment the first track, to identify a set of fifth tracks, based on one or more inflection points on the first track;

select, from sets of fourth tracks and sets of fifth tracks identified from the plurality of first tracks, a plurality of first target tracks, wherein an accuracy rate of the plurality of first target tracks is higher than an accuracy rate of a plurality of second tracks from the plurality of first tracks;

for each first target track, connect the first target track to at least one second track corresponding to the first target track, to generate a branch, wherein the branch comprises the first target track and the at least one second track corresponding to the first target track; and

generate an indoor map for indoor positioning based on a combination of the generated branches.

14 . The indoor map generation device according to claim 13 , wherein each first target track comprises a radio frequency fingerprint point that indicates a radio signal source obtained by one of the plurality of terminal devices through scanning in the process of the position movement, and a radio frequency fingerprint point comprised on the at least one second track corresponding to the first target track that matches a radio frequency fingerprint point on the first target track.

15 . The indoor map generation device according to claim 14 , wherein the at least one second track comprises a radio frequency fingerprint point matching a first radio frequency fingerprint point comprised on the first target track.

16 . The indoor map generation device according to claim 13 , wherein the program instructions, which when executed by the at least one processor, further cause the at least one processor to:

for each branch of the generated branches, correct a coordinate system of the at least one second track to a coordinate system of the first target track corresponding to the at least one second track.

17 . A non-transitory computer-readable storage medium, storing one or more instructions that, when executed by at least one processor, cause the at least one processor to:

identify a plurality of first tracks based on information received from sensors of a plurality of terminal devices within a period of time and a pedestrian dead reckoning (PDR) algorithm, wherein each first track of the plurality of first tracks represents a curve comprising a position movement of a terminal device, of the plurality of terminal devices, within the period of time, and the first track comprises a plurality of step points of the terminal device on a same floor of a building in a process of the position movement;

for each first track of the plurality of first tracks,

segment the first track, to identify a set of fourth tracks, based on a change in information from at least one of the sensors, and

segment the first track, to identify a set of fifth tracks, based on one or more inflection points on the first track;

select, from sets of fourth tracks and sets of fifth tracks identified from the plurality of first tracks, a plurality of first target tracks, wherein an accuracy rate of the plurality of first target tracks is higher than an accuracy rate of a plurality of second tracks from the plurality of first tracks;

for each first target track, connect the first target track to at least one second track corresponding to the first target track, to generate a branch, wherein the branch comprises the first target track and the at least one second track corresponding to the first target track; and

generate an indoor map for indoor positioning based on a combination of the generated branches.

18 . The non-transitory computer-readable storage medium according to claim 17 , wherein each first target track comprises a radio frequency fingerprint point that indicates a radio signal source obtained by one of the plurality of terminal devices through scanning in the process of the position movement, and a radio frequency fingerprint point comprised on the at least one second track corresponding to the first target track that matches a radio frequency fingerprint point on the first target track.

19 . The non-transitory computer-readable storage medium according to claim 18 , wherein

the at least one second track comprises a radio frequency fingerprint point matching a first radio frequency fingerprint point comprised on the first target track.

20 . The non-transitory computer-readable storage medium according to claim 17 , wherein the one or more instructions that, when executed by the at least one processor, further cause the at least one processor to:

for each branch of the generated branches, correct a coordinate system of the at least one second track to a coordinate system of the first target track corresponding to the at least one second track.