CN104850843A - Method for rapidly detecting personnel excessive gathering in high-accuracy positioning system - Google Patents
Method for rapidly detecting personnel excessive gathering in high-accuracy positioning system Download PDFInfo
- Publication number
- CN104850843A CN104850843A CN201510272508.1A CN201510272508A CN104850843A CN 104850843 A CN104850843 A CN 104850843A CN 201510272508 A CN201510272508 A CN 201510272508A CN 104850843 A CN104850843 A CN 104850843A
- Authority
- CN
- China
- Prior art keywords
- cell
- personnel
- bunch
- data points
- threshold values
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for rapidly detecting personnel excessive gathering in a high-accuracy positioning system. The method comprises the following steps: dividing a detection region according to a neightbour threshold value [lambda], projecting positioning data of n people in corresponding cells, respectively using an integer matrix to record the number of projected data points in each cell and whether a Boolean matrix labeled cell is visited; for each personnel positioning data point, using a breadth first method to visit a cell to find a cluster; calculating size of the cluster of each personnel positioning data point; when the number of data points in the cluster of a certain personnel positioning data point is larger than the set personnel threshold value, performing excessive gathering early warning prompt. The method can improve detection efficiency, reduce time complexity and can adapt to excessive detection requirements of various shapes.
Description
Technical field
The present invention relates to a kind of method that in high-accuracy position system, quick testing staff excessively assembles, a kind of method that particularly in indoor wireless locating system based on improvement CLIQUE algorithm, quick testing staff excessively assembles.
Background technology
The positioning precision of locating due to current commercial satellite is lower and be not suitable for indoor positioning environment, and the application system demand at present based on indoor wireless location or hi-Fix occurs in a large number.In these positioning systems, user often has the needs of various detection alarm to realize the real-time monitor and managment to personnel.Wherein, realize real-time personnel excessively to assemble detection alarm and be a kind of unconventional demand and there is very important practical significance.Excessive gathering is detected mainly according to the aggregation extent parameter that system manager provides, and finds possible excessive aggregation zone fast.Safety-security area and need real-time testing staff's situation densely populated place scene under, as airport, prison, factory etc., positioning system can judge personnel's concentrations region fast in real time, for avoid security incident or potential to have a fist fight, the generation of the illegal hazard event such as gathering, real-time monitoring management carried out to personnel have important function and meaning.But multiple concurrent real-time task is often run on the backstage of positioning system, the excessive early warning of such as real-time traffic, accompany access to report to the police, break in warning, card loss detection, obtain up-to-date locator data, personnel positions real-time update, position location is data-optimized with correction etc., therefore, and real-time excessive gathering test problems a kind of efficient effective and to be applicable to the detection method of complex scene very necessary.
Detecting can by directly judging whether the number in certain surface area exceedes certain threshold values to distinguish in excessive gathering (hereinafter referred to as gathering or excessively gathering).Calculate according to this mode, need the minimum encirclement convex polygon of directly trying to achieve personnel positioning data set subset, then computing staff's density, judge whether density of personnel exceedes prescribed threshold.This mode is directly perceived, but the number of the data subset that will solve exponentially increases along with the increase of personnel amount, and thus time complexity is exponential time complexity.And the area asking subset minimum encirclement convex polygon and minimum encirclement convex polygon to cover is also relatively complicated, this method also do not adopted often in the engineering of reality.
Another kind of more direct mode be in traversal surveyed area point, in the α neighborhood judging each point, whether personnel's number of (rectangle or circle shaped neighborhood region) exceedes certain threshold values to judge.This method time complexity is O (M*L*N), and wherein M*L detects in surveyed area rectangle to count, and N is locating personnel's number.This Method compare simply and easily person of being managed is understood, but it trends towards detecting round gathering, for the non-round excessive gathering (hereinafter referred to as length gathering more than 2 αs) of some aggregation zone maximum lengths more than 2 α, as strip or dendritic gathering often bad the and its time complexity height of effect depend on the number of space mid point, this is huge challenge to the server end of computation-intensive.By space fully being segmented and complexity can being reduced to O (M*L*D) by the mode of the number of matrix record each cell, D is a constant relevant with segmentation degree to Size of Neighborhood, but this mode fully will be segmented space, to make error less, this makes M*L excessive, space complexity becomes O (M*L), and this method still to there is efficiency not high, and the non-circular gathering as strip cannot be solved.
It is travel through personnel that the one improved a kind of upper method is improved one's methods, and judges whether the number in the α neighborhood of each personnel exceedes certain threshold values.This mode needs to obtain each point and institute's distance a little, then judges and this distance is less than all of α and counts, so algorithm complex becomes O (N*N).In sensing range, density of personnel is relatively little, this mode can be raised the efficiency.But his problem is be difficult to detect length excessively assemble more than 2 α equally, and efficiency declines serious when N is larger.Another of this method is not enough be exactly when the region of gathering be ring-type or arc-shaped time, this method is often invalid.
Summary of the invention
In order to overcome the above-mentioned shortcoming of prior art, the invention provides a kind of method that in high-accuracy position system, quick testing staff excessively assembles, the efficiency of detection can be improved, reduce time complexity and the excessive detection demand of various shape can be adapted to.
The technical solution adopted for the present invention to solve the technical problems is: a kind of method that in high-accuracy position system, quick testing staff excessively assembles, comprises the steps:
Suppose that length is a, width is have n personnel in the surveyed area of b,
Step one, according to neighbour's threshold values λ, cell division is carried out to surveyed area;
Step 2, the locator data of n personnel to be projected among corresponding cell, and respectively by number and the whether accessed mistake of Boolean matrices indexing unit lattice of the data point made a basket in an each cell of integer matrix record;
Step 3, the way access cell of breadth First is adopted to find bunch to each personnel's fix data points;
Step 4, calculate comprise each personnel's fix data points bunch size;
Step 5, when comprise certain personnel's fix data points bunch value be greater than personnel's threshold values N of setting time, then excessively assemble early warning.
Compared with prior art, good effect of the present invention is:
Excessively detection is assembled by the inventive method, when not carrying out considering that the dendroid of excessively dispersion is excessively assembled, the time degree of being mixed with of the algorithm number that to be O (n), n be in surveyed area.When the number of location in about 1000 scale time, the inventive method improves two orders of magnitude than the mode of direct traversal.Detection method wants uninterrupted repeatedly real time execution, and this method only needs to open up memory positioning states when detecting and starting, and repetition Real-Time Monitoring afterwards can share this memory source, so greatly reduce the resource overhead opening up internal memory.And then this method can save the computational resource of server end greatly.And the detection case that existing method is assembled strip, dendroid gathering or ring-type is undesirable, often can only be undertaken by the mode expanding the radius of neighbourhood, but often make threshold values too low like this, there is the situation a large amount of excessively gathering being detected.But the inventive method can detect the excessive gathering situation of this form, and can arrange with to not wishing that detection case is rejected neatly.
Embodiment
By carrying out combing again to the demand of excessively assembling detection, we consider, if the number in certain area is too much, then the personnel's spacing in this region is relatively near.We abstract for the positional information of personnel be point spatially.Consider based on this, we can determine personnel's set of region and the gathering assembled by the distance between personnel.Make as given a definition spatially putting us.
Definition one: definition λ is neighbour's threshold values, if the distance between two points is less than λ, is then with λ neighbour (or both assemble mutually) both claiming, is called for short 2 neighbours.Wherein, a and he itself is put with any λ neighbour.
Definition two: defining point a and some b can reach with λ, if there is series of points c
1... c
n, make a and c
1with λ neighbour, b and c
nwith λ neighbour, and c
iwith c
i+1with λ neighbour.
The subset s of definition three: one point set u, if this subset s mid point be all each other with λ can reach and in s mid point and u-s arbitrfary point can not reach with λ; the neighbour bunch of title s in u with λ being neighbour's threshold values, abbreviation s is neighbour bunch.
Definition four: the size of neighbour bunch is the number of data point in neighbour bunch.
By above definition, we define excessively to assemble and are: what there is personnel positioning data point composition is that the size of the neighbour bunch of neighbour's threshold values exceedes given threshold values with λ.In this way, the concept of personnel's Relatively centralized spatially can represent with the neighbour of composition of personnel bunch.
By illustrating above, excessively assemble the problem detecting the size that can be converted into each neighbour that data point in detection space is gathered into bunch and whether exceed threshold values.And this mode is similar to the thought of cluster.Cluster based on distance is by the distance of the data in space by definition, automatically close data point is divided into a class.This tradition generally can only process spherical class based on the clustering algorithm of distance and process data quantitative change large time, efficiency is not high.More crucially, the class that this mode is polymerized to automatically usually do not meet our needs bunch definition, need further to change, the often relative complex of conversion between the two, effect in practical application is also bad.
Another large class is based on space segmentation or the method for density.In indoor locating system, its orientation range is determined often in systems development process, and location to as if personnel, its space density can not be large especially.So density based and space divided method are a kind of feasible solutions.Wherein Clique is the algorithm of a kind of density based and space segmentation, and its thought is very close with this application, and we are applied in gathering detection by simplifying and improving a kind of new method of clique algorithm proposition.
The central idea of CLIQUE clustering algorithm is as follows: the set of (1) given multidimensional data point, and data point is not equiblibrium mass distribution usually in data space.CLIQUE distinguishes sparse in space and " crowded " region (or unit), to find the global distributed schema of data acquisition.(2) if the data point comprised in a unit has exceeded certain input model parameter, then this unit is intensive.In CLIQUE algorithm, bunch to be defined as the maximum set (herein bunch to define large neighbour before bunch different, be the set of dense cell lattice) of connected dense cell herein.According to the central idea of above-mentioned this algorithm, the clustering algorithm based on grid and density may be defined as: by data space split into grid shape, then the number of the point fallen in certain unit is treated as the density of this unit.At this moment can specify a numerical value, when the number of the various point of certain lattice unit is greater than this numerical value, just say that this cell is intensive.Finally, cluster is also just defined as the set of all " intensive " cells of connection.
Excessively assemble in test problems personnel, also can use for reference the thought according to clique, space is subdivided into the space that several are little, to look for bunch by searching close quarters, and then find corresponding neighbour bunch, then judge bunch whether to meet excessive aggregation conditions.To clique algorithm, we carry out some and improve and simplify, and reduce difficulty and the complexity of algorithm.
Because the data dimension in location is lower, all two coordinate datas, the dml algorithm carrying out complicated subspace clustering and complexity is not needed to carry out subspace beta pruning, namely accumulation unit need not be found from one-dimensional case, but directly carry out judgement dense cell by the quantity of anchor point in cell in two-dimensional space, thus enormously simplify the difficulty of algorithm.In addition, clique is that the complexity of time depends on the degree that space is segmented by carrying out cluster to the search of all space cells.But in positioning system, the quantity of personnel is often far smaller than the cell of Region dividing, and therefore this way of search efficiency is not very high.This method proposes a kind of algorithm being undertaken searching for by personnel and mark, and Flag-based clique, hereinafter referred to as fb-clique.Specific practice is: first project in cell by locator data, records the number of data in each cell.Secondly, all personnel's locator data is traveled through, to the timi requirement of each locator data o (1) in the cell at place, from this unit, search element by breadth First mode.In the process of search, need the whether accessed mistake of indexing unit lattice, mark by mark matrix record.If cell corresponding to certain data point is not accessed, cell is labeled as access, this point is put in set s.Afterwards, the cell be not labeled in 8 cells around this cell is visited; If locator data is 0 in cell, be labeled as access; If data are greater than 0, be labeled as access, join in set of data points, and from these be added into set s points further search unit lattice, until there is no new addressable cell.Finally, s is one bunch, realizes bunch time complexity searched like this to be O (n), n be the quantity of personnel.
In addition, with clique stress and strain model according to the expectation of fine degree carry out dividing unlike, in this method, the division of grid is carried out according to neighbour's threshold values λ.This needs system manager to provide the threshold values λ of personnel neighbour.When carrying out stress and strain model, according to the length of λ, two coordinate axis are divided.Divide in this manner, we can draw following reasoning:
Reasoning one: two points being less than arbitrarily λ apart necessarily drop in two adjacent unit or in same cell.
Except this, we will regulation cell be also the conditions of dense cell lattice.We can define, if having data point in the cell of spatial division, then claim this cell to be intensive.Namely bunch in addition, if two intensive cells are adjacent, then defining two cells is local denses, finds out the maximum close quarters that can search by local dense regional increment formula.
We have following reasoning:
Reasoning two: fb-clique is bunch equal by a neighbour of the set of data point in any maximum close quarters (bunch) found out and data centralization.
Prove: because the dense cell lattice length found out is λ, must mutually can reach so count in close quarters, and clique find bunch in data point and other data point not neighbour, thus data set forms a neighbour bunch.Namely bunch in like manner, a neighbour bunch of data centralization, is bound to form a close quarters in the space divided.The two one_to_one corresponding.Therefore, we are found neighbour bunch and can be undertaken by the cell of finding space segmentation bunch.Judge whether to occur gathering by the size calculating neighbour in data point set bunch can to become in computer memory cell and bunch comprise number of data points to judge.
λ is not only the tightness degree that have impact on neighbour's bunch mid point, and impact is segmented space.By the gathering regulating λ can detect different aggregation extent.Be in operation, keeper also will provide the threshold values N of bunch size, if a bunch size exceedes threshold values, is then judged to excessive gathering occurs.
Provide neighbour's threshold values, excessively assemble detection with fb-clique, then the personnel one meeting mutual near neighbor to fix in a neighbour bunch and are detected.And this method can find length more than the gathering of 2 α or tree-shaped gathering.But excessive gathering tree-shaped in the use of reality detects to be selected according to scene.Need under some scene to carry out similar detection, but some scene does not need such detection.Because dendritic excessive gathering often produce excessively dispersion bunch.The situation that straight line or the dendroid of too disperseing assemble is formed for data point, whether configuration detection can be needed according to management scene.
In practice, the situation that the dendroid that we explore needs rejecting too to disperse by following several mode process is excessively assembled, deflation approach and limited depth backtracking.Deflation approach and limited depth backtracking method can reduce by mistake alert probability.
Tighten ratio juris fairly simple, keeper, except providing neighbour's threshold values size, also additionally will provide the maximum length expecting to detect excessive aggregation zone.By fb-clique find bunch, judge whether the boundary rectangle of bunch mid point meets area size condition to judge whether excessive gathering.Specific practice is that algorithm is looked in the process of neighbour bunch and recorded neighbour bunch corresponding space bunch at x, minimum, maximal value in y-axis, and then the boundary rectangle obtaining bunch (space bunch).Number threshold values N is less than (namely if counted in bunch, do not meet excessive aggregation conditions), or bunch in count and be greater than number threshold values N and boundary rectangle size is less than the region threshold values (the region threshold values namely set) that keeper provides, then whether need not carry out detecting is the situation that the dendroid of too disperseing excessively is assembled again.If bunch count and be greater than number threshold values N and boundary rectangle size is greater than the region threshold values of setting, then exceed the adjacent boundary on the limit of the zone length of setting from boundary rectangle and cut one deck grid, the one deck cut is satisfy condition to comprise minimum one of data point in border, form new bunch and boundary rectangle, terminate when counting and be less than number threshold values N in new bunch, otherwise continue to judge whether the boundary rectangle of new bunch is greater than the region threshold values of setting: in this way, then exceed the adjacent boundary on the limit of zone length to cut from boundary rectangle and comprise the minimum one deck grid of data point, continue to form new bunch and boundary rectangle, go down successively until bunch boundary rectangle be less than the region threshold values of setting, then count and whether be still greater than given threshold values in judging bunch, then produce to assemble in this way and report to the police.
Suppose when searching close quarters, l is cell number in close quarters, cell number in the boundary rectangle of k close quarters.Close quarters forms the square of maximum to be catercorner length be l in bunch boundary rectangle as defined above, data point number on limit will be judged when carrying out boundary rectangle trimming, time complexity is directly proportional to boundary rectangle cell quantity, so the worst efficiency of algorithm is o (n
2).But in practical application, bunch due to more concentrated, and can form multiple bunches, l/k is much larger than 1/l, and av eff is similar to o (n).This method forms several comparatively intensive small sets in excessive aggregation zone inside, and these small sets close to boundary rectangle edge time, the situation of excessively assembling early warning and failing to report may be produced, thisly fail to report the performance that can not affect warning in practice.Be a kind of concept intuitively and need the experience of keeper to set to close on threshold values because excessively assemble, situation about generally failing to report can detect by regulating neighbour's threshold values and number threshold values.
Limited depth backtracking method is by carrying out breadth traversal to each data point, limits the maximum level degree of depth of traversal when traveling through.Limited by the degree of depth, can detect around a point, in certain distance, whether excessive gathering occurs, this with directly judge that the method for whether excessively assembling in the α neighborhood of each point is similar, but time complexity is improved.With certain put the region that starts to carry out searching for be judged as unduly assemble after to carry out mark and remove, with to start except this traversal some correspondence cell except access flag remove.Assuming that limited depth dp, then the worst time complexity of this algorithm is O (n*dp*dp), and n is the number of data point, and dp is generally smaller in practice, is often set between 3-7 in practice.
It is below the excessive detailed step assembling detection of carrying out of fb-clique.The dendroid not carrying out rejecting excessively dispersion in this step is excessively assembled.
Suppose that length is a, width is have n personnel in the surveyed area of b, detection algorithm step following (if need the dendroid of carrying out rejecting excessively dispersion excessively to assemble, also will provide degree of depth restriction dp or desired region length):
Step one, be a to length, width is that the surveyed area of b divides according to neighbour's threshold values λ, obtains p*q cell, wherein:
Step 2, the locator data of n personnel to be projected among corresponding cell, and respectively by the number of the data point made a basket in each cell of integer matrix record on p*q rank and the whether accessed mistake of Boolean matrices indexing unit lattice on p*q rank.
Projecting method: suppose that the coordinate data of certain personnel is for (x, y), then to coordinate be
cell in data point number add 1, initial value is 0.
Step 3, the way access cell of breadth First is adopted to find bunch to each personnel's fix data points:
(1) cell at i-th personnel's fix data points place is found;
(2) the whether accessed mistake of this cell is judged: if so, then return the cell that (1) step finds the i-th+1 personnel's fix data points place; If not, perform (3), with this cell for start element lattice carry out breadth first traversal find comprise this point bunch;
(3) start element case marker be designated as access, and start element lattice are put in set s, s deposit found bunch in element;
(4) by not accessed in cell adjacent around start element lattice and comprise in cell number of data points be 0 cell be labeled as access;
(5) by not accessed in cell adjacent around start element lattice and comprise in cell number of data points be not 0 cell be labeled as access, and put into set s, then successively using the cell of these new marks as new start element lattice;
(6) repeat (3) to (5), until there are not new start element lattice, then enter next step;
I-th personnel's fix data points that what step 4, calculation procedure three obtained comprise bunch in comprise data point.
By comprise i-th personnel's fix data points bunch in the locator data that comprises of all cells count be added obtain comprising i-th personnel's fix data points bunch in the number of data points that comprises;
Step 5, judge to comprise i-th personnel's fix data points bunch in number of data points whether be greater than threshold values N, if so, then export " around i-th personnel's fix data points excessively assemble " and also carry out early warning; Then the cell that (1) step finds the i-th+1 fix data points place is returned, until the cell at n personnel's fix data points place is all traversed.
On algorithm backstage detect thread need constantly to call execution, in order to recycle memory headroom, detect in thread at one, the storage matrix that mark matrix and representation space subdivision unit lattice are deposited can reuse, and allows and represents that the data matrix of cell and the mark matrix of indexing unit lattice are shared in each detection.Specific practice is after algorithm has run, then the algorithm calling a breadth traversal carries out mark removing.This is for fewer in number, and when space segmentation is thinner highly significant, need not detects so at every turn and all will carry out opening space.In continuous cycle detection, for once Memory Allocation of o (m*n) and initialization, knows data and mark afterwards only with o (n) at every turn.
Mark removes algorithm:
(1) cell at fix data points place is found;
(2) judge the whether accessed mistake of this cell: if not, then return the cell that (1) step finds next fix data points place; If so, perform (3), clear data with this cell for start element lattice carry out breadth first traversal and mark.
(3) start element case marker is designated as does not access, and by number of data points zero setting in this cell.
(4) by accessed in cell adjacent around this cell and comprise in cell number of data points be 0 cell be labeled as access.
(5) by accessed around start element lattice and comprise in cell count be not 0 cell be labeled as and do not access, and remove record data, the cell then these newly cleared data successively is as new start element lattice.
(6) repeat (4) to (5), return (1) after new start element lattice until can not find.
(7) until all data points are all traversed, remove mark and terminate.
Claims (6)
1. the method that in high-accuracy position system, quick testing staff excessively assembles, is characterized in that: comprise the steps:
Suppose that length is a, width is have n personnel in the surveyed area of b,
Step one, according to neighbour's threshold values λ, cell division is carried out to surveyed area;
Step 2, the locator data of n personnel to be projected among corresponding cell, and respectively by number and the whether accessed mistake of Boolean matrices indexing unit lattice of the data point made a basket in an each cell of integer matrix record;
Step 3, the way access cell of breadth First is adopted to find bunch to each personnel's fix data points;
Step 4, calculate comprise each personnel's fix data points bunch in number of data points;
Step 5, when comprise certain personnel's fix data points bunch in number of data points be greater than personnel's threshold values N of setting time, then excessively assemble early warning.
2. the method that in a kind of high-accuracy position system according to claim 1, quick testing staff excessively assembles, is characterized in that: the number of the cell that step one obtains is p*q, wherein:
3. the method that in a kind of high-accuracy position system according to claim 1, quick testing staff excessively assembles, it is characterized in that: described in step 2 to the method that the locator data of each personnel projects be: suppose that the coordinate data of certain personnel is for (x, y), then to coordinate be
cell in data point number add 1, initial value is 0.
4. the method that in a kind of high-accuracy position system according to claim 1, quick testing staff excessively assembles, is characterized in that: described in step 3 to the method that each personnel's fix data points adopts the way access cell of breadth First to find bunch be:
(1) cell at some personnel positioning data point places is found;
(2) the whether accessed mistake of this cell is judged: if so, then return the cell that (1) step finds next personnel's fix data points place; If not, with this cell for start element lattice, perform (3);
(3) start element case marker is designated as accesses, and start element lattice are put in set s;
(4) by not accessed in cell adjacent around start element lattice and comprise in cell number of data points be 0 cell be labeled as access;
(5) by not accessed in cell adjacent around start element lattice and comprise in cell number of data points be not 0 cell be labeled as access, and put into set s, then successively using the cell of these new marks as new start element lattice;
(6) repeat (3) to (5), until there are not new start element lattice, the S set obtained is bunch.
5. the method that in a kind of high-accuracy position system according to claim 1, quick testing staff excessively assembles, is characterized in that: the method for the size of compute cluster described in step 4 is: by bunch in the locator data that comprises of all cells to count the size being added and namely obtaining bunch.
6. the method that in a kind of high-accuracy position system according to claim 1, quick testing staff excessively assembles, it is characterized in that: described in step 5 when comprise certain personnel's fix data points bunch value be greater than personnel's threshold values N of setting time, judge whether the boundary rectangle of this bunch is greater than the region threshold values of setting: in this way, then exceed the adjacent boundary on the limit of zone length from boundary rectangle and cut one deck and comprise the minimum grid of data point, form new bunch and boundary rectangle, terminate this to detect when the value of new bunch is less than personnel's threshold values N of setting, otherwise go down until the boundary rectangle of new bunch is less than the region threshold values of setting successively, and then judge whether the size of new bunch is still greater than personnel's threshold values N of setting, then excessively assemble early warning in this way.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510272508.1A CN104850843B (en) | 2015-05-26 | 2015-05-26 | A kind of method that quick testing staff excessively assembles in high-accuracy position system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510272508.1A CN104850843B (en) | 2015-05-26 | 2015-05-26 | A kind of method that quick testing staff excessively assembles in high-accuracy position system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104850843A true CN104850843A (en) | 2015-08-19 |
| CN104850843B CN104850843B (en) | 2018-05-01 |
Family
ID=53850478
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510272508.1A Active CN104850843B (en) | 2015-05-26 | 2015-05-26 | A kind of method that quick testing staff excessively assembles in high-accuracy position system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104850843B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105760484A (en) * | 2016-02-17 | 2016-07-13 | 中国科学院上海高等研究院 | Crowd treading pre-warning method and system and server with system |
| CN105764029A (en) * | 2016-04-19 | 2016-07-13 | 福州市佳璞电子商务有限公司 | RFID transceiver, video positioning system based on RFID, and video positioning method based on RFID |
| CN108846438A (en) * | 2018-06-15 | 2018-11-20 | 电子科技大学 | A kind of matching process of forming a team based on real geographical location |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101464944A (en) * | 2007-12-19 | 2009-06-24 | 中国科学院自动化研究所 | Crowd density analysis method based on statistical characteristics |
| CN101751668A (en) * | 2009-11-13 | 2010-06-23 | 北京智安邦科技有限公司 | Method and device for detecting crowd density |
| US20110040758A1 (en) * | 2009-08-17 | 2011-02-17 | Cheng-Fa Tsai | Grid-based data clustering method |
| CN102509151A (en) * | 2011-11-08 | 2012-06-20 | 上海交通大学 | Video-processing-based crowd density and distribution estimation method |
| CN104268528A (en) * | 2014-09-28 | 2015-01-07 | 深圳市科松电子有限公司 | Method and device for detecting crowd gathered region |
-
2015
- 2015-05-26 CN CN201510272508.1A patent/CN104850843B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101464944A (en) * | 2007-12-19 | 2009-06-24 | 中国科学院自动化研究所 | Crowd density analysis method based on statistical characteristics |
| US20110040758A1 (en) * | 2009-08-17 | 2011-02-17 | Cheng-Fa Tsai | Grid-based data clustering method |
| CN101751668A (en) * | 2009-11-13 | 2010-06-23 | 北京智安邦科技有限公司 | Method and device for detecting crowd density |
| CN102509151A (en) * | 2011-11-08 | 2012-06-20 | 上海交通大学 | Video-processing-based crowd density and distribution estimation method |
| CN104268528A (en) * | 2014-09-28 | 2015-01-07 | 深圳市科松电子有限公司 | Method and device for detecting crowd gathered region |
Non-Patent Citations (1)
| Title |
|---|
| 陈卓 等: "一种基于网格和密度凝聚点的快速聚类算法", 《哈尔滨工业大学学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105760484A (en) * | 2016-02-17 | 2016-07-13 | 中国科学院上海高等研究院 | Crowd treading pre-warning method and system and server with system |
| CN105760484B (en) * | 2016-02-17 | 2019-10-25 | 中国科学院上海高等研究院 | A kind of crowd tramples method for early warning, system and the server with the system |
| CN105764029A (en) * | 2016-04-19 | 2016-07-13 | 福州市佳璞电子商务有限公司 | RFID transceiver, video positioning system based on RFID, and video positioning method based on RFID |
| CN108846438A (en) * | 2018-06-15 | 2018-11-20 | 电子科技大学 | A kind of matching process of forming a team based on real geographical location |
| CN108846438B (en) * | 2018-06-15 | 2022-05-24 | 电子科技大学 | Team matching method based on real geographic position |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104850843B (en) | 2018-05-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pelanis et al. | Indexing the past, present, and anticipated future positions of moving objects | |
| Potamias et al. | Sampling trajectory streams with spatiotemporal criteria | |
| Zhang et al. | Effectively indexing uncertain moving objects for predictive queries | |
| Gao et al. | Continuous visible nearest neighbor query processing in spatial databases | |
| Huang et al. | Detecting regions of disequilibrium in taxi services under uncertainty | |
| Gao et al. | Visible reverse k-nearest neighbor query processing in spatial databases | |
| Tampakis et al. | Scalable distributed subtrajectory clustering | |
| CN110070121A (en) | A kind of quick approximate k nearest neighbor method based on tree strategy with balance K mean cluster | |
| Gao et al. | Continuous visible nearest neighbor queries | |
| US9317529B2 (en) | Memory-efficient spatial histogram construction | |
| CN104850843A (en) | Method for rapidly detecting personnel excessive gathering in high-accuracy positioning system | |
| Kriegel et al. | Proximity queries in large traffic networks | |
| Chen et al. | An adaptive updating protocol for reducing moving object database workload | |
| CN110135180A (en) | Meet the degree distribution histogram dissemination method of node difference privacy | |
| Wang et al. | An approach based on maximal cliques and multi-density clustering for regional co-location pattern mining | |
| Xu et al. | An improved LOF outlier detection algorithm | |
| CN108764272A (en) | A kind of electricity consumption multi dimensional space clustering method based on Getis-Ord Gi* | |
| Rslan et al. | Spatial R-tree index based on grid division for query processing | |
| Yang et al. | Clues: a unified framework supporting interactive exploration of density-based clusters in streams | |
| Chen et al. | Towards efficient mit query in trajectory data | |
| Lv et al. | Identify and trace criminal suspects in the crowd aided by fast trajectories retrieval | |
| Wang et al. | Research of mining algorithms for uncertain spatio-temporal co-occurrence pattern | |
| Gao et al. | Processing mutual nearest neighbor queries for moving object trajectories | |
| Corral et al. | Predictive join processing between regions and moving objects | |
| Yang et al. | Distributed grid-based k nearest neighbour query processing over moving objects |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |