CN111523577A - Mass trajectory similarity calculation method based on improved LCSS algorithm - Google Patents

Mass trajectory similarity calculation method based on improved LCSS algorithm Download PDF

Info

Publication number
CN111523577A
CN111523577A CN202010285620.XA CN202010285620A CN111523577A CN 111523577 A CN111523577 A CN 111523577A CN 202010285620 A CN202010285620 A CN 202010285620A CN 111523577 A CN111523577 A CN 111523577A
Authority
CN
China
Prior art keywords
track
time
similarity
points
trajectory
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010285620.XA
Other languages
Chinese (zh)
Inventor
刘宇
耿鑫
李国栋
李维
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Fiberhome Telecommunication Technologies Co ltd
Original Assignee
Nanjing Fiberhome Telecommunication Technologies Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing Fiberhome Telecommunication Technologies Co ltd filed Critical Nanjing Fiberhome Telecommunication Technologies Co ltd
Priority to CN202010285620.XA priority Critical patent/CN111523577A/en
Publication of CN111523577A publication Critical patent/CN111523577A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/26—Visual data mining; Browsing structured data

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses a mass trajectory similarity calculation method based on an improved LCSS algorithm, which comprises the following steps: acquiring real-time position information of a moving object based on acquisition equipment to obtain track data; verifying the obtained track data; accessing the verified track data into an HBASE database; inputting a similar track and a specific date of an object track to be inquired through a web front end, and inquiring the current-day track of the object according to the object ID; and (3) issuing the current track T _0 of the object to a big data platform in a parameter form, and starting the MR task to calculate the similarity after the big data platform schedules the task. The invention improves the problem that the prior LCSS algorithm is sensitive to the selection of a time threshold when calculating the similarity of the trace points; the problem of real-time performance of track similarity calculation under the condition of a large data set is solved; the effect of mining and displaying the associated tracks of the objects is achieved.

Description

Mass trajectory similarity calculation method based on improved LCSS algorithm
Technical Field
The invention discloses a mass trajectory similarity calculation method based on an improved LCSS algorithm, and relates to the technical field of internet information.
Background
In the existing object track information positioning processing method, when the acquisition equipment acquires the position information of the moving object, the problem of uneven sampling occurs, so that when the track similarity between the objects is calculated, the similarity between the tracks needs to be matched from the perspective of global optimum.
In the conventional LCSS algorithm in the prior art, when track points are compared, the problem of sensitivity to the selection of a time difference threshold of a time-space track point occurs. Meanwhile, because the data volume of the object track is large, if the algorithm is operated on a single machine when the track similarity is calculated, the calculation speed is slow, the time required by calculation is long, and the real-time performance of the system is poor.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: aiming at the defects of the prior art, the method for calculating the similarity of the mass tracks based on the improved LCSS algorithm is provided, the similarity calculation is carried out on the track information of the moving object through a specific algorithm, and the potential space-time relation object of the object can be obtained based on the algorithm, so that the space-time relation map of the object is further improved. In the invention, the improved LCSS algorithm is defined as LCSS +, and the algorithm overcomes the problem of sensitivity to a time threshold from the point that the sum of the similarity of the tracing points is the maximum target, and is superior to the LCSS algorithm under the average condition and more accurate than the LCSS algorithm. Meanwhile, the invention utilizes the Hadoop cluster to give full play to the operational capability of the machine, uses MapReduce as a calculation frame thereof, and divides and cures calculation tasks, thereby shortening the time required by calculation, designing and realizing a distributed LCSS + algorithm, and further improving the real-time performance of the system.
The invention adopts the following technical scheme for solving the technical problems: a mass track similarity calculation method based on an improved LCSS algorithm comprises the following steps:
acquiring real-time position information of a moving object through acquisition equipment to obtain track data;
step two, checking the obtained track data, judging whether necessary fields in the obtained position information have values and have correct formats, and discarding the data if no values or the correct formats exist;
step three, the verified track data is accessed into an HBASE database;
inputting a similar track and a specific date of the track of the object to be inquired through the web front end, and inquiring the track of the object on the current day according to the object ID;
step five, carrying out the current day track T of the object0And issuing the data to a big data platform in a parameter form, and starting an MR task to calculate the similarity after the big data platform schedules the task.
Further, the specific step of starting the MR task to calculate the similarity in step five includes:
501. each track point in the object track data is defined on the object to form a complete track, and the track identity card number is taken as a key and the longitude, the latitude and the time of the object are taken as values at the display end for outputting;
502. after the track of the object is obtained, segmenting track points in time;
if a plurality of track points exist in the time period, respectively calculating the central point according to the time, the longitude and the latitude until only one track point exists in the time period to form a processed track T1；
To the track T1And track T0Performing intersection calculation on time, and respectively taking the common track points of the appearance time;
calculating the track similarity by using an LCSS + algorithm, and outputting the track similarity by taking the object ID as key and the track similarity as value;
503. and solving k objects with the similarity to the track of the specified object larger than a set threshold, and outputting a result to a storage library after the solving is finished.
Further, in the lcs + algorithm of step 502, the basic definition required for the spatio-temporal trajectory calculation specifically includes:
definition 1: the locus points in the space-time locus are composed of ternary attributes, namely time, longitude and latitude, and the ith space-time locus point in the space-time locus is marked as: pi＝(ti，logi，lati)，tiRepresents the time of the ith track point in the track, logiLongitude, lat, representing the ith trace pointiRepresenting the latitude of the ith track point;
definition 2: the space-time trajectory is composed of all trajectory points in the trajectory, and one trajectory i is recorded as: t isi＝{Pi1，Pi2，…，Pin|in＝len(Ti)}；
Definition 3: the time series in the track T is denoted ts,
Figure BDA0002448398260000021
Figure BDA0002448398260000022
definition 4: minimum time of the space-time trajectory T is recorded as
Figure BDA0002448398260000023
The maximum time of the space-time trajectory T is recorded as
Figure BDA0002448398260000024
Definition 5: space-time trajectory TaTime series ts ofaAnd a space-time trajectory TbTime series ts ofbThe intersection between is defined as: ts isa∩tsb＝{P(ti)|MaxTa≥t≥MinTa∩MaxTb≥t≥MinTb}；
Definition 6: space-time trajectory TaAnd a space-time trajectory TbThe intersection between is defined as: t isa∩Tb＝{Pi|Pi∈Ta||Pi∈Tb，Pi(t)∈(tsa∩tsb)}；
Definition 7: if the space-time trajectory TaMinimum time MinT ofaAnd maximum time MaxTaWith space-time trajectory TbMinimum time MinT ofbAnd maximum time MaxTbSatisfy MinTb≤MinTa≤MaxTa≤MaxTbThen there is Ta∈tTb。
Furthermore, the track points include longitude and latitude in position, the distance between the track points needs to be calculated when the track points are compared, and the distance between the two track points is as follows:
Figure BDA0002448398260000031
wherein lat1 represents the longitude of track point 1, llog1 represents the latitude of track point 1, and similarly lat2, log2 represents the longitude and latitude of track point 2;
if dis (P)ai(lat)，Pai(log)，Pbj(lat)，Pbj(log)) <, the two trace points are considered to be a pair of spatially similar points in space.
Further, the similarity calculation rule of the two trace points is as follows:
if the two track points are not similar in space, the two track points are not similar;
if the two track points are similar in space, the closer the two track points are in time, the more similar the two track points are, otherwise, the more dissimilar the two track points are;
two tracing points Pai、PbjThe similarity value of (d) is set as:
Figure BDA0002448398260000032
when tracing point PaiAnd locus point PbjWhen the difference in time is less than t, the similarity between the trace points is 1;
when the difference of the track points in time is larger than t, the similarity of the track points approaches to 0.
Further, when the LCSS + algorithm compares two tracks, the goal is to maximize the sum of the similarity of the track points;
definition dp [ i ]][j]Is a track TaNumbering from 1 to i-1 constitutes a subsequence, track TbThe largest similarity they contain, numbered from 1 to j-1, the constituent subsequences;
when the track TaPoint of track PaiAnd track TbRail ofLocus PbiSatisfy the requirement of
dis(Pai(lat)，Pai(log)，Tbj(lat)，Pbj(log)) < equal to or less than the total weight of the mixture,
Figure BDA0002448398260000041
when dis > dp [ i ] [ j ] ═ max { dp [ i ] [ j-1], dp [ i-1] [ j ] };
when i is 0 or j is 0, dp [ i ] [ j ] ═ 0.
As a further preferable scheme of the invention, in the step one, the real-time location information of the object includes longitude, latitude, identity ID or time; in addition, format check and data filtering are required to be performed on the data during collection.
Compared with the prior art, the invention adopting the technical scheme has the following technical effects: the invention improves the problem that the prior LCSS algorithm is sensitive to the selection of a time threshold when calculating the similarity of the trace points; the problem of real-time performance of track similarity calculation under the condition of a large data set is solved; the effect of mining and displaying the associated tracks of the objects is achieved.
Drawings
Fig. 1 is a flow chart of mass trajectory similarity calculation.
FIG. 2 is a schematic diagram of a LCSS + algorithm state transition matrix solving process;
fig. 3 is a schematic diagram of the state transition matrix solving process of the lcs algorithm.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
The technical scheme of the invention is further explained in detail by combining the attached drawings:
the invention provides a LCSS + calculation method which is more accurate in time-space trajectory similarity, and a distributed LCSS + algorithm is designed based on a Hadoop platform in massive time-space trajectory calculation, so that the real-time performance of the system is improved.
A flow chart of mass trajectory similarity calculation is shown in fig. 1, and the method includes:
the method comprises the steps that firstly, real-time position information of a mobile object is collected through collection equipment, the collection content mainly comprises longitude, latitude, identity ID, time and the like, and format verification and data filtering are carried out on data during collection.
Step two, verifying the acquired track data: and judging whether the necessary fields in the acquired information have values and are in correct format, and discarding the piece of data if no values or the correct format exists.
And step three, accessing the track data into an HBASE database.
And step four, inputting the similar track and the specific date of the track of the object to be inquired by the web front end, and inquiring the track of the current day of the object according to the ID of the object.
Step five, tracking the object track T0And issuing the data to a big data platform in a parameter form, and starting an MR task to calculate the similarity after the big data platform schedules the task.
The specific steps of calculating the similarity of the MR tasks comprise:
a) the object track points are discrete track points when stored in the database. Then the individual trace points must first be fixed to the person to form a complete trace. And outputting the object longitude, latitude and time as value by taking the track identity card number as key at the Map end.
b) In the Reduce stage, after the track of the object is obtained, the track points are segmented in time, the time period can be set through a configuration file, if a plurality of track points exist in the time period, the central points are respectively obtained aiming at time, longitude and latitude, and finally only one track point exists in the track point in the time period to form the processed track T1Subsequently to the track T1And track T0Performing intersection operation on time, respectively taking the common track points of the appearance time, then calculating the track similarity by using an LCSS + algorithm, and finally calculating the track similarityAnd outputting the object ID as a key and the track similarity as a value.
c) And the second MapReduce process mainly solves k objects with higher similarity to the track of the specified object, and mainly solves the problem of topK and outputs a result to a storage library after the solution is finished.
In step b), the LCSS + algorithm defines the basic definition required for the spatio-temporal trajectory calculation due to the particularity of the spatio-temporal trajectory and the description of the subsequent algorithm, and includes the following steps:
definition 1: the locus points in the space-time locus are composed of ternary attributes, namely time, longitude and latitude, and the ith space-time locus point in the space-time locus is marked as: pi＝(ti，logi，lati)，tiRepresents the time of the ith track point in the track, logiLongitude, lat, representing the ith trace pointiRepresenting the latitude of the ith trace point.
Definition 2: the space-time trajectory is composed of all trajectory points in the trajectory, and one trajectory i is recorded as: t isi＝{Pi1，Pi2，…，Pin|in＝len(Ti)}
Definition 3: the time series in the track T is denoted ts,
Figure BDA0002448398260000061
Figure BDA0002448398260000062
definition 4: minimum time of the space-time trajectory T is recorded as
Figure BDA0002448398260000063
The maximum time of the space-time trajectory T is recorded as
Figure BDA0002448398260000064
Definition 5: space-time trajectory TaTime series ts ofaAnd a space-time trajectory TbTime series ts ofbThe intersection between is defined as: ts isa∩tsb＝{P(ti)|MaxTa≥t≥MinTa∩MaxTb≥t≥MinTb}
Definition 6: space-time trajectory TaAnd a space-time trajectory TbThe intersection between is defined as: t isa∩Tb＝{Pi|Pi∈Ta||Pi∈Tb，Pi(t)∈(tsa∩tsb)}
Definition 7: if the space-time trajectory TaMinimum time MinT ofaAnd maximum time MaxTaWith space-time trajectory TbMinimum time MinT ofbAnd maximum time MaxTbSatisfy MinTb≤MinTa≤MaxTa≤MaxTbThen there is Ta∈tTb。
Because track point includes longitude and latitude only in the position, need calculate its distance apart when comparing track point, the distance of two track points is:
Figure BDA0002448398260000065
where lat1 represents the longitude of track point 1, llog1 represents the latitude of track point 1, and similarly lat2, log2 represents the longitude and latitude of track point 2.
If dis (P)ai(lat)，Pai(log)，Pbj(lat)，Pbj(log)) <, the two trace points are considered to be a pair of spatially similar points in space.
Because the time attribute is required to be kept at the time-space track point, the similarity calculation rule of the two track points is defined as follows: if neither of the two trace points are spatially similar, then the two trace points are not similar. If two trace points are similar in space, the closer they are in time, the more similar they are, otherwise the more dissimilar they are.
Then two tracing points Pai、PbjThe similarity value of (d) may be set as:
Figure BDA0002448398260000071
when tracing point PaiAnd locus point PbjWhen the difference in time is less than t, the similarity between the track points is 1, and when the difference in time of the track points is greater than t, the similarity of the track points can be close to 0, so that the method accords with the practical situation.
When two tracks are compared by the LCSS + algorithm, the goal is to maximize the sum of the similarity of the track points, the main idea of the LCSS + algorithm is to calculate the sum of the similarity of the track points by the maximum goal, the problem that the LCSS + is also a dynamic plan can be found, and dp [ i ] is defined][j]Is a track TaNumbering from l to i-l constituting a subsequence, track TbThe subsequences are numbered from l to j-l, which contain the largest sum of similarities.
When the track TaPoint of track PaiAnd track TbPoint of track PbiSatisfy the requirement of
dis(Pai(lat)，Pai(log)，Pbj(lat)，Pbj(log)) < equal to or less than the total weight of the mixture,
Figure BDA0002448398260000072
when dis > dp [ i ] [ j ] ═ max { dp [ i ] [ j-1], dp [ i-1] [ j ] };
when i is 0 or j is 0, dp [ i ] [ j ] ═ 0.
The pseudo-code of the algorithm is as follows:
LCSS+(Ta，Tb)
1 lena←len(Ta)
2 lenb←len(Tb)
3dp[][]←new dp[lena][lenb]
4 for i←0 to lena
5dp[i][0]←0
6 forj←0 to lenb
7dp[0][j]←0
8 for i←1 to lena
9 for j←1 to lenb
10Pj←Tb(j) // j-th node of trace b
11Pi←Ta(i)
12 if dis≤
13
Figure BDA0002448398260000081
dp[i][j-1]，dp[i-1][j]}，
14 elsedp[i][j]←max{dp[i][j-1]，dp[i-1][j]}
15 return dp[lena][lenb]/(math.min{lena，lenb})。
Pseudo code 4-7 line time complexity is: theta (n), with 8 to 14 lines of temporal complexity theta (n)2) The average time complexity of the algorithm is theta (n)2)。
In order to distinguish the proposed LCSS + algorithm from the conventional LCSS calculation method for measuring the similarity of space-time trajectories, a specific embodiment is given below to illustrate the problem of sensitivity to time difference threshold selection when the conventional LCSS algorithm compares trajectory points.
As shown in FIG. 2, the arrow pair indicates the transition of the cumulative similarity sum in the case that the two spatial trajectories are similar in spatial location, for example, if B (9: 32) and B (9: 50) belong to point B in spatial location and differ in time by 19 minutes, then its cumulative optimal similarity sum is:
Figure BDA0002448398260000082
dp[2][2]，dp[1][3]the maximum value between B (9: 40) and B (10: 02) is calculated to be 1.22, and is spatially related to point B, with a temporal difference of 22 minutes, then from dp [2 ]][3]The cumulative maximum similarity sum of the transitions is: 1/(1+2.2 × 2.2) +1.22 is 1.39. From dp [3 ]][3]The cumulative maximum similarity sum of the transferred sequences is 1.45, obviously greater, so dp 3][4]Should be 1.45. The similarity is 2.45/5-0.49. Although two tracks can take two public points B at the point B in the process of traveling, the time difference of track points positioned at the two public points B is larger and is not as large as the similarity accumulation brought by one public point B, namely the point B ((9: 40) and the point B (9: 51).
Referring to the state transition matrix of the corresponding LCSS algorithm in the solving process, as shown in FIG. 3, when the LCSS algorithm solves the state matrix, there are only two common similarity points, the first pair of similarity points is A (9: 25) and A (9: 30), and the second pair of similarity points is D (10: 13) and D (10: 15). The finally solved maximum similarity sum is 2, the similarity is 0.4, compared with the LCSS + algorithm, the problem of time interval sensitivity is highlighted, when 10 minutes are selected as interval points, the LCSS algorithm may lose some possibly similar common points, such as B (9: 40) and B (9: 51), because the common points are cut off after 11 minutes of difference, so that some precision is lost in the track similarity comparison, and finally the common points may not appear in the relation of the objects.
The distributed lcs + algorithm is mainly embodied in the design of the MR program, and now a program pseudo code is given:
Figure BDA0002448398260000083
Figure BDA0002448398260000091
Figure BDA0002448398260000101
the pseudo code of the Reduce phase is similar to the Map phase and will not be repeated. And outputting the objects with the maximum k track similarity values through the two dependent MapReduce tasks.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of the ordinary skilled person in the art without departing from the gist of the present invention. Although the present invention has been described with reference to the above embodiments, it should be understood that the invention is not limited to the above embodiments, and various changes and modifications can be made by the above embodiments without departing from the scope of the invention.

Claims (7)

1. A mass trajectory similarity calculation method based on an improved LCSS algorithm is characterized by comprising the following steps:
acquiring real-time position information of an object through acquisition equipment to obtain track data;
step two, checking the obtained track data, judging whether necessary fields in the obtained position information have values and have correct formats, and discarding the data if no values or the correct formats exist;
step three, the verified track data is accessed into an HBASE database;
inputting a similar track and a specific date of the track of the object to be inquired through the web front end, and inquiring the track of the object on the current day according to the object ID;
step five, carrying out the current day track T of the object0And issuing the data to a big data platform in a parameter form, and starting an MR task to calculate the similarity after the big data platform schedules the task.
2. The method for calculating the similarity of the mass trajectories based on the improved lcs algorithm as claimed in claim 1, wherein the step five of starting the MR task to calculate the similarity comprises the steps of:
501. each track point in the object track data is defined on the object to form a complete track, and the track identity card number is taken as a key and the longitude, the latitude and the time of the object are taken as values at the display end for outputting;
502. after the track of the object is obtained, segmenting track points in time;
if a plurality of track points exist in the time period, aiming at the time and the longitudeRespectively calculating the central point of the degree and the latitude until only one track point is formed in a time period to form a processed track T1；
To the track T1And track T0Performing intersection calculation on time, and respectively taking the common track points of the appearance time;
calculating the track similarity by using an LCSS + algorithm, and outputting the track similarity by taking the object ID as key and the track similarity as value;
503. and solving k objects with the similarity to the track of the specified object larger than a set threshold, and outputting a result to a storage library after the solving is finished.
3. The method for calculating the similarity of mass trajectories based on the improved lcs s algorithm as claimed in claim 2, wherein the basic definition of the spatio-temporal trajectory calculation required in the lcs s + algorithm of step 502 specifically includes:
definition 1: the track points in the space-time track are composed of ternary attributes, namely time, longitude and latitude, and the ith space-time track point in the space-time track is marked as Pi＝(ti,logi,lati)，tiRepresents the time of the ith track point in the track, logiLongitude, lat, representing the ith trace pointiRepresenting the latitude of the ith track point;
definition 2 the space-time trajectory is composed of all trajectory points in the trajectory, and one trajectory i is marked as Ti＝{Pi1,Pi2,…,Pin|in＝len(Ti)}；
Definition 3: the time series in the track T is denoted ts,
Figure FDA0002448398250000021
Figure FDA0002448398250000022
definition 4: minimum time of the space-time trajectory T is recorded as
Figure FDA0002448398250000023
The maximum time of the space-time trajectory T is recorded as
Figure FDA0002448398250000024
Definition 5: space-time trajectory TaTime series ts ofaAnd a space-time trajectory TbTime series ts ofbThe intersection between them is defined as tsa∩tsb＝{P(ti)|MaxTa≥t≥MinTa∩MaxTb≥t≥MinTb}；
Definition 6: space-time trajectory TaAnd a space-time trajectory TbThe intersection between them is defined as Ta∩Tb＝{Pi|Pi∈Ta||Pi∈Tb,Pi(t)∈(tsa∩tsb)}；
Definition 7: if the space-time trajectory TaMinimum time MinT ofaAnd maximum time MaxTaWith space-time trajectory TbMinimum time MinT ofbAnd maximum time MaxTbSatisfy MinTb≤MibTa≤MaxTa≤MaxTbThen there is Ta∈tTb。
4. The method for calculating the similarity of mass trajectories based on the improved lcs s algorithm as claimed in claim 2, wherein the positions of the trajectory points in step 502 include longitude and latitude, the distance between the trajectory points needs to be calculated when comparing the trajectory points, and the distance between two trajectory points is:
Figure FDA0002448398250000025
wherein lat1 represents the longitude of track point 1, llog1 represents the latitude of track point 1, and similarly lat2, log2 represents the longitude and latitude of track point 2;
if dis (P)ai(lat),Pai(log),Pbj(lat),Pbj(log))<Then the two trace points are considered to be a pair of spatially similar points in space.
5. The method for calculating the similarity of mass trajectories based on the improved LCSS algorithm as claimed in claim 4, wherein the similarity calculation rule of two trajectory points is:
if the two track points are not similar in space, the two track points are not similar;
if the two track points are similar in space, the closer the two track points are in time, the more similar the two track points are, otherwise, the more dissimilar the two track points are;
two tracing points Pai、PbjThe similarity value of (d) is set as:
Figure FDA0002448398250000031
when tracing point PaiAnd locus point PbjWhen the difference in time is less than t, the similarity between the trace points is 1;
when the difference of the track points in time is larger than t, the similarity of the track points approaches to 0.
6. The method for calculating the similarity of mass trajectories based on the improved LCSS algorithm as claimed in claim 4, wherein: when the LCSS + algorithm compares two tracks, the goal is to maximize the sum of the similarity of the track points;
definition dp [ i ]][j]Is a track TaNumbering from 1 to i-1 constitutes a subsequence, track TbThe largest similarity they contain, numbered from 1 to j-1, the constituent subsequences;
when the track TaPoint of track PaiAnd track TbPoint of track PbiSatisfy the requirement of
dis(Pai(lat),Pai(log),Pbj(lat),Pbj(log)) < equal to or less than the total weight of the mixture,
Figure FDA0002448398250000032
when dis > dp [ i ] [ j ] ═ max { dp [ i ] [ j-1], dp [ i-1] [ j ] };
when i is 0 or j is 0, dp [ i ] [ j ] ═ 0.
7. The method for calculating similarity of mass trajectories based on improved LCSS algorithm as claimed in claim 1, wherein in the first step, the real-time location information of the object includes longitude, latitude, ID or time; and carrying out format check and data filtering on the data during acquisition.
CN202010285620.XA 2020-04-13 2020-04-13 Mass trajectory similarity calculation method based on improved LCSS algorithm Pending CN111523577A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010285620.XA CN111523577A (en) 2020-04-13 2020-04-13 Mass trajectory similarity calculation method based on improved LCSS algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010285620.XA CN111523577A (en) 2020-04-13 2020-04-13 Mass trajectory similarity calculation method based on improved LCSS algorithm

Publications (1)

Publication Number Publication Date
CN111523577A true CN111523577A (en) 2020-08-11

Family

ID=71902935

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010285620.XA Pending CN111523577A (en) 2020-04-13 2020-04-13 Mass trajectory similarity calculation method based on improved LCSS algorithm

Country Status (1)

Country Link
CN (1) CN111523577A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111950937A (en) * 2020-09-01 2020-11-17 上海海事大学 Key personnel risk assessment method based on fusion space-time trajectory
CN112131278A (en) * 2020-09-28 2020-12-25 浙江大华技术股份有限公司 Method and device for processing track data, storage medium and electronic device
CN112434084A (en) * 2020-12-02 2021-03-02 电信科学技术第十研究所有限公司 Trajectory similarity matching method and device based on geohash and LCSS
CN113126132A (en) * 2021-04-09 2021-07-16 内蒙古科电数据服务有限公司 Method and system for calibrating and analyzing track in mobile inspection
CN113487865A (en) * 2021-07-02 2021-10-08 江西锦路科技开发有限公司 System and method for acquiring information of vehicles running on highway
CN113806463A (en) * 2021-09-06 2021-12-17 李莉 Trajectory similarity calculation method based on space-time pyramid matching
CN115795115A (en) * 2023-02-11 2023-03-14 云南师范大学 Multi-track set similarity searching method based on graph storage

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6812756B2 (en) * 2001-10-16 2004-11-02 Altera Corporation PLL/DLL circuitry programmable for high bandwidth and low bandwidth applications
CN110505583A (en) * 2019-07-23 2019-11-26 中山大学 A kind of path matching algorithm based on bayonet data and signaling data
CN110728842A (en) * 2019-10-23 2020-01-24 江苏智通交通科技有限公司 Abnormal driving early warning method based on reasonable driving range of vehicles at intersection
CN110941278A (en) * 2019-12-20 2020-03-31 交控科技股份有限公司 In-station dynamic security analysis method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6812756B2 (en) * 2001-10-16 2004-11-02 Altera Corporation PLL/DLL circuitry programmable for high bandwidth and low bandwidth applications
CN110505583A (en) * 2019-07-23 2019-11-26 中山大学 A kind of path matching algorithm based on bayonet data and signaling data
CN110728842A (en) * 2019-10-23 2020-01-24 江苏智通交通科技有限公司 Abnormal driving early warning method based on reasonable driving range of vehicles at intersection
CN110941278A (en) * 2019-12-20 2020-03-31 交控科技股份有限公司 In-station dynamic security analysis method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
涂刚凯: "海量时空轨迹相似度计算方法研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111950937A (en) * 2020-09-01 2020-11-17 上海海事大学 Key personnel risk assessment method based on fusion space-time trajectory
CN111950937B (en) * 2020-09-01 2023-12-01 上海海事大学 Important personnel risk assessment method based on fusion of space-time trajectories
CN112131278A (en) * 2020-09-28 2020-12-25 浙江大华技术股份有限公司 Method and device for processing track data, storage medium and electronic device
CN112434084A (en) * 2020-12-02 2021-03-02 电信科学技术第十研究所有限公司 Trajectory similarity matching method and device based on geohash and LCSS
CN112434084B (en) * 2020-12-02 2023-04-07 电信科学技术第十研究所有限公司 Trajectory similarity matching method and device based on geohash and LCSS
CN113126132A (en) * 2021-04-09 2021-07-16 内蒙古科电数据服务有限公司 Method and system for calibrating and analyzing track in mobile inspection
CN113487865A (en) * 2021-07-02 2021-10-08 江西锦路科技开发有限公司 System and method for acquiring information of vehicles running on highway
CN113487865B (en) * 2021-07-02 2022-07-22 江西锦路科技开发有限公司 System and method for acquiring information of vehicles running on highway
CN113806463A (en) * 2021-09-06 2021-12-17 李莉 Trajectory similarity calculation method based on space-time pyramid matching
CN113806463B (en) * 2021-09-06 2023-04-28 北京信息科技大学 Track similarity calculation method based on space-time pyramid matching
CN115795115A (en) * 2023-02-11 2023-03-14 云南师范大学 Multi-track set similarity searching method based on graph storage

Similar Documents

Publication Publication Date Title
CN111523577A (en) Mass trajectory similarity calculation method based on improved LCSS algorithm
CN107103754B (en) Road traffic condition prediction method and system
US20200210890A1 (en) Trajectory cluster model for learning trajectory patterns in video data
Yan et al. SeTraStream: semantic-aware trajectory construction over streaming movement data
CN110874362A (en) Data association analysis method and device
CN109241349B (en) Monitoring video multi-target classification retrieval method and system based on deep learning
TWI425454B (en) Method, system and computer program product for reconstructing moving path of vehicle
CN109634946B (en) Intelligent track matching correlation analysis method based on big data mining
CN109543312B (en) Space-time investigation analysis method and system
CN112734809B (en) On-line multi-pedestrian tracking method and device based on Deep-Sort tracking framework
CN103080975A (en) Face image registration device and method
US11829192B2 (en) Method, apparatus, and computer program product for change detection based on digital signatures
CN112035591A (en) Road network matching method, device, equipment and storage medium
CN109740479A (en) A kind of vehicle recognition methods, device, equipment and readable storage medium storing program for executing again
JP2022511093A (en) Device message framework
US20220318555A1 (en) Action recognition using limited data
EP4209959A1 (en) Target identification method and apparatus, and electronic device
CN110727740B (en) Correlation analysis method and device, computer equipment and readable medium
CN114265833A (en) Vehicle track guiding geographic space-time characteristic visualization analysis platform
CN112364176A (en) Method, equipment and system for constructing personnel action track
WO2021114615A1 (en) Method, apparatus, and device for visualization of behavior risk identification, and storage medium
CN111709382A (en) Human body trajectory processing method and device, computer storage medium and electronic equipment
Ding et al. Mit-avt clustered driving scene dataset: Evaluating perception systems in real-world naturalistic driving scenarios
CN113313098A (en) Video processing method, device, system and storage medium
CN106682168B (en) Construction method of cross-domain urban data visual query system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20200811