CN108537241A - A kind of building moving object track method for measuring similarity - Google Patents
A kind of building moving object track method for measuring similarity Download PDFInfo
- Publication number
- CN108537241A CN108537241A CN201710119129.8A CN201710119129A CN108537241A CN 108537241 A CN108537241 A CN 108537241A CN 201710119129 A CN201710119129 A CN 201710119129A CN 108537241 A CN108537241 A CN 108537241A
- Authority
- CN
- China
- Prior art keywords
- track
- semantic
- distance
- point
- similarity
- 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
Links
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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/42—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
- G06V10/422—Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation for representing the structure of the pattern or shape of an object therefor
- G06V10/424—Syntactic representation, e.g. by using alphabets or grammars
Abstract
The present invention provides a kind of building moving object track method for measuring similarity, belong to mobile object management technical field in CyberSpace database.Computational methods based on traditional trajectory distance, the computational methods for proposing the track of the semantic information fusion of interior space structure and distance and position solve the problems, such as that existing trajectory distance computational methods are not directly applicable the interior space and the huge similitude of calculation amount.Motion track reconstruction is carried out using the characteristic locus data of big clutter extraction, greatly reduces computation complexity.Simultaneously, it will expand to the similarity measurement of semantic locations track, consider space, these three factors of time to the corresponding position of semantic track similarity measurement Position Design semantic relation tree and extraction of semantics algorithm, in conjunction with the distance calculating method of the three-dimensional track of indoor environment, indoor track is handled by the data normalization of comprehensive similarity.
Description
Technical field
The present invention relates to a kind of building moving object track method for measuring similarity, belong in CyberSpace database
Moving object administrative skill field.
Background technology
Currently, the similarity calculation of movement locus is concentrated mainly on the exterior space or road network space.Euclidean distance quilt
For calculating the similitude of the movement locus in the exterior space.It is calculated first by the coordinate of Euclidean distance, then to entire
The weighted average similarity of track, this method calculate simply, it can be readily appreciated that but only hardly resulting in reality to the track of equal length
It applies on border.DTW（Dynamic time warping distance）The value of time coordinate point reach the dynamic time before 2 path integration of local time
Regular clone method, the length that can calculate errant is different, but this method is to noise-sensitive.M. Vlachos is based on
The limitation of LCS method longest common subsequences solution path length, and the similarity degree of the orbital segment to vary in weight, effectively solve
The certainly track of noise problem.The result shows that this method is better than Euclidean distance and classical way DTW.Woods et al. is by defining distance
Distance function measure the similitude of movement locus shape.
The above method only considered the similitude of space and shape, not consider influence of the time factor to similitude.
Pelekis et al. considers time factor, it is proposed that the method for measuring similarity of the exterior space is a method, they propose
The concept of lip, on the basis of the spatial simlanty that some small regional two Orbital Overlaps surround calculates, while with two tracks and time
Similarity consideration maximum time area ratio, considering for influence factor time and space obtain final similitude.
Track is considered the combination of multi-line section by skoumas et al., in the period of two tracks corresponding horizontal distance, space, away from
From the composite factor calculating with vertical angle, the synthesis similitude of space tracking.
In addition to the distance of the room and time of the similitude used and the track of shape are come the track similitude measured, also have
The influence of the track similarity measurement of one important position.Ying et al. increases position in the similarity calculation of track
The method that the space cell of semantic information, position and semantic marker divides is added, and then the method for border on the sea warship calculates two rails
The similitude in road.For building space with the exterior space there are significant difference, the above method is not directly applicable building space.Mesh
Before, it is also fewer about building track similarity measurements quantifier elimination.Health et al. has been put forward for the first time a kind of based on border on the sea war 2009
The computational methods of track analogue method in bucket warship, by two tracking statistics of the time locus similitude in overlapping in same position
It sets, the method proposed is simple, but excessively coarse for trajectory calculation, and semantic location information is only answered by character
It is hardly resulted in substitution.Then, Wang et al. utilizes Euclidean distance and editing distance method, overcomes only lacking using position as character
Point.Gold et al. expression is compared with geographic coordinate information, and the more positions of semantic information are by similarity value, they devise one
A concept hierarchy Move Mode is for measuring the semantic information between building position, to obtain the position preference of user.
Invention content
Goal of the invention
The purpose of the present invention is to propose to a kind of building moving object track method for measuring similarity.Due to being directed to moving object at present
Track similarity measurements quantifier elimination is based on the exterior space more, less to the research of building space, however building space and outdoor are empty
Between structure it is different, moving object positioning method is different, the factors such as motion track dimension difference, existing to be moved based on the exterior space
The inquiry research of dynamic rail mark is not directly applicable building space.It is asked to solve the motion track similarity query under building environment
Topic, needs to define new three-dimensional track distance metric method according to building space structure and motion track feature.Building environment
The difference of moving object track not only exists only in time and space, focuses more on the difference of position semantic information, thus by when
Empty and position semantic information is included in similarity measurement and is calculated simultaneously has realistic meaning with the different demands for adapting to different user.
Technical solution
The present invention relates to a kind of measurement methods of building moving object similitude：
Step 1：A kind of track reconstructing algorithm is proposed, and defines corresponding position deviation and angular deviation threshold value, simplifies guarantor
Demonstrate,prove the integrality based on huge lengthy and jumbled track track data.
Step 2：The space-time similarity calculation of independent track：Trajectory distance, which calculates, uses three-dimensional track shadow casting technique, to sky
Between trajectory calculation formula provide distance：SLIP (T, R) = ∑ Area i *ω i *h i .WhereinArea, ωi,hIt is mobile respectively
Track projects intersecting area, weight coefficient, motion track projector distance.The time gap calculating of movement locus is divided into two kinds of sides
Formula：The intersection point of time and the intersection point of time.
Step 3：For the position of the semantic extension of motion track similarity calculation, semantic association lsr_tree tree constructions come
Pass between the position of relational design between descriptive semantics position and the semantic distance height of the orbital position of tree node definition
System, design semantic is apart from extraction algorithm.
Second step, the TR of each movement locus, according to feature point extraction algorithm, we are set as generation in available position
The characteristic point of tableCPs={cp 1 ,…, cp m }.In characteristic point sequence, each two adjacent characteristic point presses the priority group of arrival time
At an orbit segmentL, thus this track can be expressed as multiple3DThe oriented sequence of line segment composition.Assuming that path line segment table
It is shown asT={L 1 ,L 2 …,L m },R={L 1 ,L 2 …,L n }, trackT’For trackTIn trackRThe projected footprint of place two dimensional surface, note
Intersection of locus point isI={I 1 ,I 2 ,…,I q }.The closed polygon that intersection of locus point and track characteristic point are surrounded after definition projection
Interior standoff height maximum value is height distance, is denoted ash.The closing that intersection of locus point and track characteristic point are surrounded after definition projection
The area of polygon is intersection of locus area, is denoted asArea.DefinitionLength T WithLength R Motion track is indicated respectivelyTWithR's
Path length, i.e. track characteristic point number,Length T (I i , I i+1 )WithLength R (I i , I i+1 )Indicate that closing is polygon respectively
ShapeArea i Interior motion trackTWithRPath length, then weight coefficient ωi, then weight coefficient is equal to closed polygon path length
Spend the ratio of route track length.
Third walks, and in wp_tree structures, the corresponding physical location of semantic attribute that leaf node represents is in building space
Position, i.e. semantic label.Nonleaf node indicates the position classification of lower node, and rank is higher, and position is bigger, position semantic similarity
It is smaller.On the contrary, similitude is bigger.In the tree, the leaf node accessed by two feature point trajectories finds minimum public father section
Point, the height and wp_tree height of the public father of node level is than between the semantic feature point position of distance.
The semantic distance of 4th step, obtained measurement result tracking time and spatial position is at three aspects, according to different use
The needs at family, according to the entire track for the Spatial Semantics similarity that the three of different proportion factors can obtain in the following manner
's.In view of different unit and the order of magnitude, using normalization（Minimum max）Method unifies differential data, to initial data
Carry out linear transformation.
In above-mentioned steps four, in LSR_Tree structures, the leaf node accessed by two track characteristic points is sought
Its minimum public father's node is found, and defines public father's node place the ratio between level height and LSR_Tree height and is characterized
Position semantic distance between point.It is T={ P by the motion track of Based on Feature Points1,P2…,Pm}, R={P1,P2…,Pn, change
Into dynamic time wrapping algorithm：dist TR(T, R)=f(m, n)。
Advantageous effect
The present invention has the advantages that using above technical scheme is compared with the prior art：
1) present invention combines building space structure, and giving motion track reconstruct using trajectory angle offset and position offset calculates
Method, calculation amount is excessive during solving the problems, such as to calculate track similarity.
2) moving object track time-space matrix separate computations are utilized building track feature, take rail by the present invention well
Mark projection strategy, solves the problems, such as that existing trajectory distance computational methods can not directly apply to building space.
3) present invention devises moving object track position semantic association tree construction and the semantic relation of track is described,
In conjunction with tree construction hierarchical relationship and improve dynamic time warping algorithm, give moving object position Semantic Similarity Measurement side
Method solves text similarity measurement algorithm directly applying to the excessive defect of track Semantic Similarity Measurement cost.
4) present invention employs data normalizations, and the track similarity calculation data of different number grade are normalized, and give
Go out and merged space, the moving object Path Generation similarity of time and position semantic three kinds of influence factors preferably adapts to use
Family demand.
Description of the drawings
Fig. 1 is the building moving object trajectory reconstruction schematic diagram of the present invention.
Fig. 2 is that the building moving object trajectory range distance of the present invention calculates schematic diagram.
Fig. 3 is that the building moving object trajectory time distance of the present invention calculates schematic diagram.
Specific implementation mode
Technical scheme of the present invention is described in further details below in conjunction with drawings and examples, motion track similitude
Measurement is a pith in motion track analysis field, refers to finding a kind of suitable distance method to measure two
Similarity degree between motion track, including trajectory shape, space length, trajectory time etc..Motion track similarity measurement exists
There are many important practical applications in location based service.It is more for moving object track similarity measurements quantifier elimination at present
Less to the research of building space based on the exterior space, since building space is different with the structure of the exterior space, moving object is fixed
Position mode is different, the factors such as motion track dimension difference, and existing studied based on the inquiry of exterior space motion track cannot be direct
Applied to building space.Limitation of the building environment due to factors such as room, corridor, stair so that Euclidean distance, road network distance etc.
It is no longer applicable in, it is necessary to redefine the three-dimensional track space length measure for meeting building space constraint.Moreover, in building sky
Between in, the time gap and position semantic distance of motion track have a major impact track similarity measurement.As shown in Fig. 3,
Motion trackTWithRSpatial form move towards very much like, but consider time attribute, just cannot be considered similar track again.It passes
Mostly by semantic information symbolism in the trajectory analysis research of system, but in building environment, position semantic information application demand is higher,
Such as in large airport, people are common to be described as " I am in boarding gate " rather than " I am in position 1 ", so spatial form phase
As track be likely to occur prodigious difference in the angle of position semanteme.Sometimes the inquiry needs of user can give priority to, such as
The similar track for inquiring certain time period, then time similarity is with regard to relatively important.The present invention is semantic by space, time, position
It measures respectively, user can need according to oneself to being measured emphatically in terms of wherein one or more, to better adapt to difference
Demand.
As shown in Fig. 1, the track that TR is building moving object, angle offset threshold value are definedω, positional distance threshold
Valued, ifp i.label =" elevator " or " staircase ", orD θpi >ω, orD pi >d, thenp i It is a characteristic point, the beginning of track
Point and end point are classified as characteristic point automatically.Assuming thatp i To label vectorp c p c+1 Minimum euclidean distance beD pi , definitionD pi Forp i
Position offset distance.Ifp i Withp c p c+1 AngleαWithβIt is acute angle or there are right angle, then minimum range is defined as vertical line
Sectiond 1 IfαWithβIn with the presence of obtuse angle, then minimum range is apart from shorter one.Angle offset distance thresholdω：The steering of track
Angle reflects the movement tendency of track,P i P i+1 For label vector,P i For the next sampled point in track, defined herein vectorP i P i+1
With vectorP i P i+1 AngleθSupplementary angle be angle offset distance.As shown in Fig. 2, intersection of locus point and rail after definition projection
Standoff height maximum value is height distance in the closed polygon that mark characteristic point is surrounded, and is denoted ash.Intersection of locus after definition projection
The area for the closed polygon that point and track characteristic point are surrounded is intersection of locus area, is denoted asArea.Define motion trackTWithRSpace similaritySLIPFor：SLIP (T, R) = ∑ Area i *ω i *h i , wherein ωi∈ [0,1] is each enclosed area
DomainArea i Weight coefficient.As shown in Fig. 3, trackTWithRIt is handed over without the time within the period of track starting points and end point
Collection, then time gap is the absolute value of first segment track starting point and second segment track end point difference.If there is intersection, the time
Distance is ratio of the first segment track starting point to second segment track end point time difference and two intersection of locus periods.
With reference to the accompanying drawings with the further details of the embodiment of technical scheme of the present invention, motion track similarity measurement neck
One important component of the motion track analysis in domain is the suitable method of similarity distance for finding to measure between track,
Including trajectory shape, space length, time locus etc..The similarity measurement of motion track is in many of location based service
Important practical application.It is less currently based on the multiple mobile object track similarity measurements quantifier elimination of the building space exterior space,
Because of the structure of building space and the exterior space, Moving objects location, different factors are such as tieed up track, are based in different ways
Existing outdoor space track inquiry is not directly applicable building space.Due to the limitation of the factors such as building, corridor, stair,
The Euclidean distance and distance of road network are no longer applicable in.In addition, in the interior space, the time gap of movement locus and position semanteme away from
From track similarity measurement have a major impact.As shown in figure 3, the space tracking shape of T and R is closely similar, but in view of when
Between attribute, it is not construed as similar track.The analysis of traditional track of semantic information symbol, but in building environment,
In the application of the semantic information of high position, such as large airport, people usually said " I am on doorway " rather than " my position exists
1, so trajectory range shape similarity possibly is present at the semantic huge difference in angle position.Sometimes the inquiry of user needs
It concentrates, the similar track of a period is such as inquired, then time similarity is important.The present invention can measure sky respectively
Between, time and position it is semantic, user can according to oneself need weigh one or more aspects, to better adapt to difference
Demand.
As shown in Figure 1, TR's is defined as a tracking building moving object, angle offset threshold valueω, positional distance threshold
Valued, ifp i.label =" elevator " or " staircase ", orD θpi >ω, orD pi >d, thenp i It is a characteristic point, the beginning of track
Point and end point are classified as characteristic point automatically.Assuming thatp i To label vectorp c p c+1 Minimum euclidean distance beD pi , definitionD pi Forp i
Position offset distance.Ifp i Withp c p c+1 AngleαWithβIt is acute angle or there are right angle, then minimum range is defined as vertical line
Sectiond 1 IfαWithβIn with the presence of obtuse angle, then minimum range is apart from shorter one.Angle offset distance thresholdω：The steering of track
Angle reflects the movement tendency of track,P i P i+1 For label vector,P i For the next sampled point in track, defined herein vectorP i P i+1
With vectorP i P i+1 AngleθSupplementary angle be angle offset distance.As shown in Fig. 2, intersection of locus point and rail after definition projection
Standoff height maximum value is height distance in the closed polygon that mark characteristic point is surrounded, and is denoted ash.Intersection of locus after definition projection
The area for the closed polygon that point and track characteristic point are surrounded is intersection of locus area, is denoted asArea.Define motion trackTWithRSpace similaritySLIPFor：SLIP (T, R) = ∑ Area i *ω i *h i , wherein ωi∈ [0,1] is each enclosed area
DomainArea i Weight coefficient.As shown in figure 3, the starting point of the tracks T and R and the time of terminal in no time in first time rail
The intersection point distance of mark, difference of the track to absolute value point and the second endpoint trace.If there is intersection point, then time interval is the first rail
Road starts the ratio with the second track start time difference and two trajectory times.
Example 1
The embodiment of the present invention one describes building moving object track data reconstructing method, and specific steps include:
A. the sampled point number of motion track T is calculated, and initializes reconstruct track R；
B. first sampled point is directly denoted as characteristic point, is put into reconstruct track R, is denoted as a label vector；
C. Rule of judgment is defined as using track characteristic point to searchp 2 ~p n-1 Between characteristic point；
D. gained characteristic point is included in reconstruct track R, sequentially moves down label vector；
E. the last one sampled point characteristic point is directly denoted as to be put into reconstruct track R.
Example 2
The embodiment of the present invention two introduces building moving object trajectory range distance calculating method, is as follows：
A. motion track is calculatedTWithRFeature point number be path lengthlength T Withlength R ；
B. motion track crosspoint is found, all intersection points are put into set I；
C. the length of track between each two adjacent point in I is calculated separatelylength(I i , I i+1 )；
D. polygonal region area is calculatedArea i ；
E. the projection maximum height h of polygonal region is calculated；
F. by track segment lengthlength(I i , I i+1 )It is weights omega to add up and take ratio with total trajectory lengthi；
G. by polygonal region area, project maximum height and weight take product after return.
Example 3
Shown in the calculating track position semantic distance of the embodiment of the present invention three is as follows：
A, semantic tree height high, minimum public father node are initialized；
B, motion track is obtainedTWithRFeature point number be path lengthlength T Withlength R ；
C, position semantic classification trees are traversed；
D, minimum public father node between two nodes of calculating, obtains its height h；
E, it calculatesh/highAs the position semantic similarity between two characteristic points of motion trackdist(pi,pj)；
F, willdist(pi,pj)It is combined with improved dynamic time warping algorithm；
G, return movement object space semantic similarityD sem (T,R)。
Claims (1)
1. the present invention relates to a kind of measurement methods of building moving object similitude：
Step 1：A kind of track reconstructing algorithm is proposed, and defines corresponding position deviation and angular deviation threshold value, simplifies guarantor
The integrality based on huge lengthy and jumbled track track data is demonstrate,proved,
Step 2：The space-time similarity calculation of independent track：Trajectory distance, which calculates, uses three-dimensional track shadow casting technique, to sky
Between trajectory calculation formula provide distance：SLIP (T, R) = ∑ Area i *ω i *h i ,
WhereinArea, ωi,hBe respectively motion track projection intersecting area, weight coefficient, motion track projector distance,
The time gap calculating of movement locus is divided into two ways：The intersection point of time and the intersection point of time,
Step 3：For the position of the semantic extension of motion track similarity calculation, semantic association lsr_tree tree constructions describe
Relationship between the position of relational design between semantic locations and the semantic distance height of the orbital position of tree node definition,
Design semantic apart from extraction algorithm,
Step 4：The standard method of range data will not commensurate, the normalization of different size of trajectory distance, initial data reflects
The data being mapped between section [0,1], the space and time order synthesis for obtaining motion track is similar,
The first step first defines the movement locus of characteristic point：The building moving object Trajectory Design of TR, angle offset threshold valueω, positional distance threshold valued,Ifp i.label =" elevator " or " staircase ", orD θpi >d, orD pi >d, thenp i It is one
The starting points and end point of characteristic point, track is classified as characteristic point automatically,
The marker bit being connect between feature vector and sampled point that minimum Eustachian distance offset distance threshold value D passes through adjacent orbit
Point, label vector and sampled point joint angle calculate the angle between income offset distance threshold value ω,
Second step, the TR of each movement locus, according to feature point extraction algorithm, our available positions are set as representative
Characteristic pointCPs={cp 1 ,…, cp m },
In characteristic point sequence, each two adjacent characteristic point forms an orbit segment by the priority of arrival timeL, thus this rail
Mark can be expressed as multiple3DThe oriented sequence of line segment composition,
Assuming that track line segment is expressed asT={L 1 ,L 2 …,L m },R={L 1 ,L 2 …,L n }, trackT’For trackTIn trackRPlace two
The projected footprint of dimensional plane, note intersection of locus point areI={I 1 ,I 2 ,…,I q },
Standoff height maximum value is height in the closed polygon that intersection of locus point and track characteristic point are surrounded after definition projection
Distance is denoted ash,
The area for the closed polygon that intersection of locus point and track characteristic point are surrounded after definition projection is intersection of locus area, note
ForArea,
DefinitionLength T WithLength R Motion track is indicated respectivelyTWithRPath length, i.e. track characteristic point number,Length T (I i , I i+1 )WithLength R (I i , I i+1 )Closed polygon is indicated respectivelyArea i Interior motion trackTWithRTrack
Length, then weight coefficient ωi, then weight coefficient be equal to closed polygon path length route track length ratio,
Third walks, in wp_tree structures, position of the corresponding physical location of semantic attribute that leaf node represents in building space
It sets, i.e. semantic label,
Nonleaf node indicates the position classification of lower node, and rank is higher, and position is bigger, and position semantic similarity is smaller,
On the contrary, similitude is bigger,
In the tree, the leaf node accessed by two feature point trajectories finds minimum public father node, the public father's of node level
Height and wp_tree height than between the semantic feature point position of distance,
The semantic distance of 4th step, obtained measurement result tracking time and spatial position is at three aspects, according to different use
The needs at family, according to the entire track for the Spatial Semantics similarity that the three of different proportion factors can obtain in the following manner
,
In view of different unit and the order of magnitude, using normalization（Minimum max）Method unifies differential data, to initial data
Carry out linear transformation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710119129.8A CN108537241A (en) | 2017-03-02 | 2017-03-02 | A kind of building moving object track method for measuring similarity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710119129.8A CN108537241A (en) | 2017-03-02 | 2017-03-02 | A kind of building moving object track method for measuring similarity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108537241A true CN108537241A (en) | 2018-09-14 |
Family
ID=63488743
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710119129.8A Pending CN108537241A (en) | 2017-03-02 | 2017-03-02 | A kind of building moving object track method for measuring similarity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108537241A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109614933A (en) * | 2018-12-11 | 2019-04-12 | 闽江学院 | A kind of motion segmentation method based on certainty fitting |
| CN109858517A (en) * | 2018-12-25 | 2019-06-07 | 中国石油大学(华东) | A kind of with the direction of motion is leading track method for measuring similarity |
| CN110196891A (en) * | 2018-11-15 | 2019-09-03 | 腾讯大地通途（北京）科技有限公司 | Determination method, apparatus, storage medium and the electronic device of type of block |
| CN110796494A (en) * | 2019-10-30 | 2020-02-14 | 北京爱笔科技有限公司 | Passenger group identification method and device |
| CN111414444A (en) * | 2020-03-02 | 2020-07-14 | 北京明略软件系统有限公司 | Method and device for processing space-time trajectory information |
| CN111783739A (en) * | 2020-07-29 | 2020-10-16 | 中国人民解放军国防科技大学 | Communication radiation source similar motion trajectory comparison method |
| CN112561948A (en) * | 2020-12-22 | 2021-03-26 | 中国联合网络通信集团有限公司 | Method, device and storage medium for recognizing accompanying track based on space-time track |
| CN112766228A (en) * | 2021-02-07 | 2021-05-07 | 深圳前海中电慧安科技有限公司 | Face information extraction method, person searching method, system, device and medium |
| CN114428807A (en) * | 2022-01-24 | 2022-05-03 | 中国电子科技集团公司第五十四研究所 | Ground maneuvering target motion trajectory semantic system construction and cognitive optimization method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334845A (en) * | 2007-06-27 | 2008-12-31 | 中国科学院自动化研究所 | Video frequency behaviors recognition method based on track sequence analysis and rule induction |
| CN103593361A (en) * | 2012-08-14 | 2014-02-19 | 中国科学院沈阳自动化研究所 | Movement space-time trajectory analysis method in sense network environment |
| CN104657424A (en) * | 2015-01-21 | 2015-05-27 | 段炼 | Clustering method for interest point tracks under multiple temporal and spatial characteristic fusion |
| CN104765873A (en) * | 2015-04-24 | 2015-07-08 | 百度在线网络技术（北京）有限公司 | Method and device for determining similarity among users |
| CN106383868A (en) * | 2016-09-05 | 2017-02-08 | 电子科技大学 | Road network-based spatio-temporal trajectory clustering method |
-
2017
- 2017-03-02 CN CN201710119129.8A patent/CN108537241A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101334845A (en) * | 2007-06-27 | 2008-12-31 | 中国科学院自动化研究所 | Video frequency behaviors recognition method based on track sequence analysis and rule induction |
| CN103593361A (en) * | 2012-08-14 | 2014-02-19 | 中国科学院沈阳自动化研究所 | Movement space-time trajectory analysis method in sense network environment |
| CN104657424A (en) * | 2015-01-21 | 2015-05-27 | 段炼 | Clustering method for interest point tracks under multiple temporal and spatial characteristic fusion |
| CN104765873A (en) * | 2015-04-24 | 2015-07-08 | 百度在线网络技术（北京）有限公司 | Method and device for determining similarity among users |
| CN106383868A (en) * | 2016-09-05 | 2017-02-08 | 电子科技大学 | Road network-based spatio-temporal trajectory clustering method |
Non-Patent Citations (3)
| Title |
|---|
| NIKOS PELEKIS 等: "Similarity Search in Trajectory Databased", 《14TH INTERNATIONAL SYMPOSIUM ON TEMPORAL REPRESENTATION AND REASONING(TIME"07)》 * |
| PEIQUAN JIN 等: "Effective Similarity Search on Indoor Moving-Object Trajectories", 《LECTURE NOTES IN COMPUTER SCIENCE》 * |
| 王倩: "室内移动对象轨迹相似性度量与应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110196891B (en) * | 2018-11-15 | 2024-03-15 | 腾讯大地通途（北京）科技有限公司 | Method and device for determining block type, storage medium and electronic device |
| CN110196891A (en) * | 2018-11-15 | 2019-09-03 | 腾讯大地通途（北京）科技有限公司 | Determination method, apparatus, storage medium and the electronic device of type of block |
| CN109614933B (en) * | 2018-12-11 | 2020-05-05 | 闽江学院 | Motion segmentation method based on deterministic fitting |
| CN109614933A (en) * | 2018-12-11 | 2019-04-12 | 闽江学院 | A kind of motion segmentation method based on certainty fitting |
| CN109858517A (en) * | 2018-12-25 | 2019-06-07 | 中国石油大学(华东) | A kind of with the direction of motion is leading track method for measuring similarity |
| CN110796494A (en) * | 2019-10-30 | 2020-02-14 | 北京爱笔科技有限公司 | Passenger group identification method and device |
| CN111414444A (en) * | 2020-03-02 | 2020-07-14 | 北京明略软件系统有限公司 | Method and device for processing space-time trajectory information |
| CN111783739A (en) * | 2020-07-29 | 2020-10-16 | 中国人民解放军国防科技大学 | Communication radiation source similar motion trajectory comparison method |
| CN112561948B (en) * | 2020-12-22 | 2023-11-21 | 中国联合网络通信集团有限公司 | Space-time trajectory-based accompanying trajectory recognition method, device and storage medium |
| CN112561948A (en) * | 2020-12-22 | 2021-03-26 | 中国联合网络通信集团有限公司 | Method, device and storage medium for recognizing accompanying track based on space-time track |
| CN112766228A (en) * | 2021-02-07 | 2021-05-07 | 深圳前海中电慧安科技有限公司 | Face information extraction method, person searching method, system, device and medium |
| CN114428807A (en) * | 2022-01-24 | 2022-05-03 | 中国电子科技集团公司第五十四研究所 | Ground maneuvering target motion trajectory semantic system construction and cognitive optimization method |
| CN114428807B (en) * | 2022-01-24 | 2023-11-03 | 中国电子科技集团公司第五十四研究所 | Method for constructing semantic system and cognition optimization of ground maneuvering target motion trail |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537241A (en) | A kind of building moving object track method for measuring similarity | |
| Teo et al. | BIM-oriented indoor network model for indoor and outdoor combined route planning | |
| Jovanović | The application of GIS and its components in tourism | |
| CN106441292B (en) | A kind of building indoor plane figure method for building up based on crowdsourcing IMU inertial guidance data | |
| McKeown | The role of artificial intelligence in the integration of remotely sensed data with geographic information systems | |
| CN102004963B (en) | Digital city generation method and system | |
| Over et al. | Generating web-based 3D City Models from OpenStreetMap: The current situation in Germany | |
| CN107491070A (en) | A kind of method for planning path for mobile robot and device | |
| CN102201129B (en) | Three-dimensional visual operation maintenance system and method of subway comprehensive pipelines | |
| Chalermpong | Rail transit and residential land use in developing countries: hedonic study of residential property prices in Bangkok, Thailand | |
| CN106462624A (en) | Tile-based geocoder | |
| CN107067038A (en) | A kind of indoor building profile modifications method drawn based on PDR tracks | |
| CN109726594A (en) | A kind of novel track data dissemination method based on difference privacy | |
| CN108108831A (en) | A kind of destination Forecasting Methodology and device | |
| CN110057362A (en) | The method for planning path for mobile robot of finite elements map | |
| Elias | Pedestrian Navigation-Creating a tailored geodatabase for routing | |
| CN108022006A (en) | The accessibility probability and Area generation method of a kind of data-driven | |
| CN110648019B (en) | Improved space syntax-based small-sized civil facility site selection method | |
| CN106339985B (en) | A method of selection is inlayed line and is inlayed to aviation image from vector house data | |
| CN113721254A (en) | Vehicle positioning method based on road fingerprint space incidence matrix | |
| Nuhn et al. | Generation of landmarks from 3D city models and OSM data | |
| CN109189871A (en) | A kind of method and apparatus of Indoor environment path planning | |
| CN108647880A (en) | A kind of water front space planning Rationality Assessment method | |
| CN104850657A (en) | Holographic position map superposing method | |
| Bieda et al. | Use of 3D technology in underground tourism: example of Rzeszow (Poland) and Lviv (Ukraine). |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180914 |
|
| WD01 | Invention patent application deemed withdrawn after publication |