CN105785411B - A kind of abnormal track-detecting method based on region division - Google Patents
A kind of abnormal track-detecting method based on region division Download PDFInfo
- Publication number
- CN105785411B CN105785411B CN201610102351.2A CN201610102351A CN105785411B CN 105785411 B CN105785411 B CN 105785411B CN 201610102351 A CN201610102351 A CN 201610102351A CN 105785411 B CN105785411 B CN 105785411B
- Authority
- CN
- China
- Prior art keywords
- track
- abnormal
- territory element
- region
- check
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/38—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system
- G01S19/39—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system the satellite radio beacon positioning system transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/42—Determining position
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The present invention proposes a kind of abnormal track-detecting method based on region division, includes classifying to the historical track of mobile object, is then divided to the region where normal trace data；Territory element extension process is carried out to the track after region division；To track regions division to be detected and extension process；Inquiring in normal trace with track to be detected there is the track of identical initiation region unit and termination area unit to gather, using supporting rate of each compositing area unit in track to be detected in normal trace set is detected, the territory element with low supporting rate enters in abnormal area unit set；The relationship for comparing the quantity of abnormal area unit set and the compositing area element number of track in normal trace set judges the abnormal conditions of track to be detected, then decides whether further to carry out subdivided detection to track regions.The present invention has carried out the subdivided detection in region according to the actual conditions of track, improves Detection accuracy and efficiency.
Description
Technical field
The invention belongs to mobile internet technical fields, relate to the use of machine learning algorithm to GPS track in mobile application
The abnormal conditions of data carry out analyzing processing, and in particular to a kind of abnormal track-detecting method based on region division.
Background technology
In recent years, the technologies such as satellite communication, GPS device, RFID, wireless sensor, Internet of Things Network Communication, video tracking monitoring
Continuous development and extensive use so that the mobile application of all size in global range is all more accurately positioned and is had
Effect tracking.By these technologies, signal receiver can collect the track data of a large amount of mobile subscribers from positioning terminal,
These data contain very abundant information, and over time, and data volume can become more and more huger, complicated,
The mass data of collection is further expanded there is an urgent need to researcher and is flexibly analyzed.
Definition according to Douglas M.Hawkins to abnormal point：One is observed a little too much with far from other, so that
It is considered the observation point that another different method is generated.Therefore the purpose of abnormal track detection is to detect and major part
The different track in track.The historical track that can be generated according to mobile subscriber itself to the abnormality detection of track is classified, and is chosen
Standard of the wherein normal track as detection is selected, that is, according to the historical movement path of analysis mobile subscriber oneself, inspection
Measure the abnormal track different from most of track；The track that can also be generated according to One-male unit user carries out abnormal track inspection
It surveys, according to the historical track that analysis user group generates, detects the single mobile subscriber that there are different tracks with user group
Generated track.In the life of reality, two methods have a scene of each self application, such as urban transportation, logistics transportation etc. by
Mobile subscriber field is limited, the movement locus of mobile subscriber is preset mostly, and abnormal track is exactly that mobile subscriber deviates from
Preset normal trace；Air particle clouds motion, animal migrate, and the track of the untethered mobile subscriber such as personal movement is then not
It is preset, normal trace library can be established according to its historical trajectory data, then by detected track and history rail
Mark is compared, and the track that historical trajectory data arrival is considered then abnormal to a certain degree is deviateed.It is used in different scenes different
Method be more of practical significance to detect in real life.The abnormal motion majority of mobile subscriber be it is unexpected, can
Huge economic loss can be caused, or even the security of the lives and property of people can be threatened.In order to preferably analyze mobile subscriber's
Active characteristics hold the activity trend of mobile subscriber and its feature of environment, must just be carried out to the track data of mobile subscriber
System is effectively analyzed and is excavated.
How to utilize and analyze the data of these huge and complicated mobile applications becomes a disaster of ongoing research area
Topic, while being also a big hot spot of research.There are numerous researchers to be carried out for user's GPS track data of acquisition at present abnormal
Detection, method substantially have：Statistics-Based Method, the method based on distance, the method based on density and the side based on depth
Method.These methods suffer from respective disadvantage and advantage, can also detect the track of user's exception to a certain extent, but
It is that these researchs come with some shortcomings.(1) it directly handles mostly and miscellaneous data, while causing the loss of track characteristic data,
It is not also guaranteed in efficiency and the accuracy rate of detection；(2) do not make detection algorithm can not according to the actual conditions of track
It is more efficiently and time saving.
Invention content
Given this object of the present invention is to provide a kind of abnormal track-detecting method based on region division, main thought
Four steps can be substantially divided into：Classified according to the historical trajectory data of target (untethered mobile application), summarizes positive normal practice
Then the feature of mark carries out region division processing, due to GPS to normal trace and track data to be detected on geographical location
The influence of data sampling frequency and mobile object movement speed is extended to dividing processed track data, so that
More and complexity data become simple and do not lose necessary feature；Then traverse track to be detected territory element set and
Normal trace territory element set, the abnormal conditions of comparison domain unit detect the abnormal conditions of track from local feature；Root
Determine whether track is abnormal track according to abnormal area unit set and normal trace set track regions cell-average length,
And point out the sub-trajectory that track is abnormal；Finally, according to the different characteristic of track, it is proposed that the subdivided abnormal track in region
Detection so that detection efficiency higher is provided to the user and more efficient, accurate is preferably serviced with real-time.
The present invention adopts the following technical scheme that achieve the goals above：A kind of abnormal track detection based on region division
Method, characterized in that include the following steps：
Step 1：The classification that normal trace data and abnormal track data are carried out to the GPS historical trajectory datas of user, carries
The rail track feature for taking normal trace data will carry out region division in normal trace data in the ground position, obtain track data
Region.It is described extraction normal trace data rail track feature include the longitude of track, track latitude and timestamp.
Step 2：The ready-portioned track data region of step 1 is subjected to track data territory element extension, obtains history
Track regions unit sequence library.
Step 3：Region division is carried out to track data to be detected and obtains unit sequence tr in track regions to be measuredcheck=
{g1,g2,...,gn, wherein gnIndicate the territory element that track to be detected is passed through, n indicates territory element serial number, from history rail
It is starting point that unit identical with the initiation region unit of track regions unit sequence to be measured is found in mark territory element sequence library, with
The identical unit of termination area unit of track regions unit sequence to be measured is terminal, composition normal trace set TR={ tr1,
tr2,tr3,...,tri, wherein i is the number of track in TR | TR |, triIndicate i-th track.tri={ gi1,gi2,
gi3,...,gij, gijIndicate track triJ-th of the territory element passed through.
Step 4：Traverse unit sequence tr in track regions to be detectedcheckIn each unit lattice track, obtain each cell
Supporting rate in normal trace set TR, and compared with threshold value, obtain abnormal area unit set A (gi)。
Step 5：According to normal trace set TR and abnormal area unit set A (gi) length relation judge it is to be detected
The abnormal conditions of track.
In order to reduce the energy consumption of detection and reduce the time of detection, the invention also includes the GPS historical track numbers to user
According to the step of carrying out subdivided region and abnormality detection with track data to be detected.
Specifically, the track data territory element extension includes that the territory element adjacent with track data region is returned
It receives into track data region, obtains historical track territory element sequence library.
In order to preferably implement the present invention, track abnormality detection is specially in the step 4：
S41：Traverse trcheckEach compositing area list, with normal trace set TR={ tr1,tr2,tr3,...,tri}
In track tri={ gi1,gi2,gi3,...,gijTerritory element compare, if track triWith trcheckArea having the same
Domain unit, then by track triIt records and deposits into track set Inc (TR, trcheck)。
S42：For each territory element gi of track to be detected, calculate each territory element Inc (TR,
trcheck) proportion function of the tracking quantity in normal trace set TR,According to proportion
Function judges whether this territory element is track trcheckIn normal trace point, as Sup (Tr, trcheck)<θ, θ are threshold value, 0
This territory element is put into abnormal area unit set A (g by≤θ≤1i) in, and work as Sup (Tr, trcheck) >=θ is by this region
Unit is put into normal region unit set N (gi) in.
It is described to judge that the abnormal conditions method of track to be detected is in above scheme：
Obtain abnormal area unit set A (gi) after, when the quantity of abnormal area unit | A (gi) | it is default more than one
ValueWhen, i.e. abnormal area unit set A (gi) number of elements | A (gi) | when more, which is abnormal；Conversely, the track
It is normal.
To the abnormal area unit set A (gi) carry out it is abnormal when judging, using calculating exceptional value R (trcheck) side
Formula judged, exceptional valueWherein α is constants, and k is and normal trace set TR
The relevant constant of average value of the territory element number of each track.
The present invention will carry out region division processing to track data and then carry out abnormality detection to the part of track, avoid
Cause the local anomaly feature probably ignored in the global property of track using entire track as basic unit
Situation, the abnormal sub-trajectory for detecting a track are more significant in practical applications；Meanwhile method is according to the practical feelings of track
Condition has carried out the subdivided detection in region, improves Detection accuracy and efficiency.
The present invention has carried out region division to track, and the input of track Outlier Detection Algorithm is territory element, to track original
The processing of beginning GPS data point is converted to the processing to track regions unit, and the time for greatly reducing track abnormality detection is complicated
Degree, it is more efficient to keep detection time shorter；The present invention can not only be detected the abnormal conditions of track entirety,
It can detect the sub-trajectory being abnormal；Furthermore present invention uses multi-level region partitioning methods, according to track original number
According to concrete condition more careful division and comprehensive detection are carried out to track, reduce algorithm to the omission factor of abnormal track and right
The false drop rate of normal trace.
Description of the drawings
The above-mentioned and/or additional aspect and advantage of the present invention, in conjunction with following accompanying drawings to that will become in the description of embodiment
Obviously and it is readily appreciated that, wherein：
Fig. 1 is the overall flow structural schematic diagram of the present invention；
Fig. 2 is that track of the present invention carries out region division schematic diagram；
Fig. 3 is region division detection algorithm flow chart of the present invention.
Specific implementation mode
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar meaning.The embodiments described below with reference to the accompanying drawings are exemplary,
It is only used for explaining the present invention, and is not considered as limiting the invention.
As shown in Figure 1, the present invention provides a kind of abnormal track-detecting method based on region division, including：To movement
The GPS historical trajectory datas of application are classified, and feature is extracted, at the trajectory map that region division is carried out to track region
Reason；Then to carry out region division after track data carry out territory element extension, make up GPS track data sample frequency and
Track characteristic caused by mobile object movement speed otherness extracts error；To track data to be detected carry out region division and from
Initiation region unit therewith and the identical track data set TR of termination area unit are extracted in normal trace sequence library；Traversal
The territory element track of trajectory map to be detected obtains supporting rate of each territory element in normal trace set, and and threshold
Value is compared, and abnormal area unit set A (g are obtainedi)；The abnormal conditions for judging track to be detected, according to the track in set TR
With set A (gi) length | A (gi) | relationship judges the abnormal conditions of track to be detected, and it is different to carry out the track that region divides again
Often detection.It is as follows：
S1：Classify to the GPS historical trajectory datas of mobile subscriber, normal trace data track characteristic is extracted, to carrying
The track of feature is taken to carry out region division mapping processing.
S2：Track data territory element extension is carried out to treated the tracks step S1, corrects the sampling of GPS track data
Track characteristic caused by the otherness of frequency and mobile object speed extracts error, obtains algorithm data input.
S3：Track data to be detected is carried out region division processing and extracted from normal trace sequence library to originate therewith
Territory element and the identical track data of termination area unit, obtain normal trace set TR.
S4：The cell track obtained after track regions to be detected divide is traversed, obtains each cell in normal trace
Supporting rate in set, and compared with threshold value, obtain abnormal area unit set A (gi)；
S5：The abnormal conditions for judging track to be detected, according to set TR and set A (gi) length relation judge it is to be detected
The abnormal conditions of track.
S6：Carry out that region is subdivided and abnormality detection according to the concrete condition of GPS track data.
The present invention is to carry out abnormal detection to the GPS track data that mobile terminal acquires, not only can be to entire track
Abnormal conditions differentiated, and may indicate that the sub-trajectory that is abnormal of track；And it can be according to the specific feelings of track
Condition to carry out the subdivided abnormality detection in region to track, and due to the subdivided abnormality detection in region, efficiency of algorithm has obtained one
Determine the raising of degree, detection time is shorter, more efficient.More there is application value in practice.
The sample of the present invention is the historical track of mobile object, classifies to GPS track historical data, extracts track
Longitude, latitude and timestamp, pi(longitude, latitude, time), piIndicate the GPS raw data points of motion track,
It is comprising three longitude, latitude and timestamp basic information.Then track collection area range is divided into specified size
Territory element scans for the range of each territory element, if there is tracing point piSo this territory element is rail
The component part of track after mark region division, if not having tracing point piThen the cell is not then the component part of track；
GPS track data become the set Tr=being made of territory element after having carried out region division processing to track data
(gi)。
The present invention carried out extension process to the territory element lattice after division, due to the sample frequency of GPS data
Reason, even the same track, region division may obtain different territory element set later, carry out cell inspection
Survey when, may be abnormal track by wherein normal track detection, testing result and actual conditions be caused not to meet.For
Solution such case, needs to being extended for track after region division, and territory element adjacent thereto is also concluded and arrives rail
In mark territory element set, timestamp is the same.
The present invention is based on the abnormal track-detecting methods of region division, extract identical initiation region unit and termination area list
The normal trace set of member：According to the track to be detected of input, initiation region unit and the end of the track are obtained after region division
Only territory element, then from normal trace sequence library to beginning and end identical track group, by these tracks, group forms
Track set TR.Then normal trace set TR={ tr have been obtained1,tr2,tr3,...,tri, wherein i is of track in TR
Number | TR |, TR and track tr to be detectedcheck={ g1,g2,g3,...,giInput as detection algorithm.
The present invention is based on the abnormal track-detecting method of region division, track abnormality detection stage etch：
S41：To trcheck={ g1,g2,g3,...,giCarry out abnormal conditions detection when, traverse trcheckEach of composition area
Domain cell, with normal trace set TR={ tr1,tr2,tr3,...,triIn track tri={ gi1,gi2,gi3,...,
gijTerritory element compare, traverse trcheckIn each territory element, TR={ tr1,tr2,tr3,...,triIt whether there is rail
Mark triWith trcheckTerritory element having the same.If there is such track tri, then by track triDeposit into track set
Inc(TR,trcheck) in, whereinTrack to be detected is calculated to reflect
Penetrate the quantity of each later territory element track identical with track regions unit in normal trace set.Gather track
Inc(TR,trcheck) weigh track tr to be detectedcheckWith normal trace set TR (identical initiation region unit and terminator
Domain unit) territory element giThe case where coincidence.
S42：For each compositing area unit g of track to be detectediHave a corresponding set Inc (TR,
trcheck), calculate Inc (TR, the tr of each territory elementcheck) tracking quantity | Inc (TR, trcheck) | in normal trace
The proportion accounted in set TR, while a threshold θ (0≤θ≤1) is defined to judge the abnormal to track of proportion according to actual conditions
Effect.Proportion functionIt can judge whether this territory element is track according to proportion function
trcheckIn normal trace point, as Sup (Tr, trcheck)<This territory element is put into abnormal area unit set A (g by θi)
In, and work as Sup (Tr, trcheckThis territory element is deposited into normal region unit set N (g by) >=θi) in, which is
The normal segments of track.Namely when the territory element of track obtain the support of more track when be considered as that track is normal
Part, however the less part for obtaining normal trace support is considered as unusual part, it, can after completion is all detected in whole track
To obtain abnormal area unit set A (gi) and normal region unit set N (gi), wherein i is A (gi) in territory element set
Number | A (gi)|。
The present invention is based on the abnormal track-detecting method of region division, judge that the abnormal conditions method of track to be detected is：
S51：Obtain abnormal area unit set A (gi) after, the detection of the abnormal conditions of track has been converted to pair
The abnormal area element number problem for forming the territory element of path, when the quantity of abnormal area unit | A (gi) | it is more than
One preset valueWhen,Value be usually the 1/2 of track to be detected composition territory element quantity, i.e. abnormal area unit collection
Close A (gi) number of elements | A (gi) | when more, which is abnormal；Conversely, the track is normal.
S52：Exceptional value determining type is as follows：Wherein α is constants, and k is and set
The relevant constant of average value of the territory element number of each tracks TR.The judgement of track abnormal conditions is not only needed to detect different
The number of normal territory element | A (gi) |, the also average length of the track of normal trace set is more in line with actual feelings in this way
Condition avoids judging by accident.A kind of following special situation can also be avoided according further to the average length of the track of normal trace set,
The accuracy higher for the detection for being.
The further optimizing detection algorithm of the present invention, when GPS track carries out region division mapping, the length of side of territory element is
It is changeless, also just say that the size of territory element is constant, the size of territory element is determined according to the type of GPS data,
Prodigious deviation will not occur for the result of detection.But it is smaller to work as territory element, forms the territory element opposed area unit of track
Larger amt is more, and detection algorithm can cause more energy consumption and time loss to the processing of territory element, can be to the property of system
It can require also higher.The present invention proposes a kind of based on the subdivided detection method in region, reduction detection according to the actual conditions of track
Energy consumption and reduce detection time.The basic thought of method is to be detected first to track using larger territory element,
In this way can in the case where territory element is larger can detected the obvious track of some off-notes, be not required to
The smaller processing of territory element is carried out, the territory element of the track of composition increases, the calculation amount of the algorithm needs of abnormality detection
Increase, the increase of calculation amount, the requirement to the time also increases accordingly, if it is online detection, then will result in timeliness
The bad effect of property.The subdivided abnormal track detection algorithm in region that the present invention uses will reduce system to a certain extent
It consumes energy and the timeliness of detection is made to be improved to some extent.
It is different since the tracing point fallen in identical strip path curve falls the data volume in same territory element, adopts
Sample frequency is fixed, and speed is then various, can cause the different (distance=speed/frequency of the distance that mobile object moves
Rate), i.e. the distance between mobile object track data point difference, so the region after the present invention is used to carrying out region division
Unit track is extended, and traverses all data point in track first, obtain trajectory map area planar map cell
Sequence, then in order to reduce error, (including itself one shares nine for the adjacent area unit of the territory element fallen into tracing point
It is a) be also required to assign a basic numerical value (territory element for being denoted as 1 in Fig. 2), prevent the identical track in path generate compared with
Big otherness, makes testing result and actual conditions there are larger deviation, and track abnormal conditions are judged by accident.Grey in Fig. 2
Territory element is the territory element that falls into of track data point, and value is the quantitative value for the data point for falling into the territory element, in figure
Use giIt indicates；The value of its neighbours' territory element is 1；The value of other territory elements is 0.Track can be expressed as tr=(g1,
g2,...,g11) wherein further include territory element sequence neighbours' territory element.
Fig. 3 is the flow chart of invention algorithm, and the input of algorithm first is the historical trajectory data of mobile object and to be detected
Track data carries out region division and extension to two class data respectively and obtains historical track territory element sequence library TR and to be detected
The territory element sequence tr of trackcheck={ g1,g2,...,gn}；Then by TR and trcheck={ g1,g2,...,gnCalculate
trcheckIn each territory element proportion function Sup values, the magnitude relationship of Sup and threshold value is judged, if territory element
Sup values are smaller than threshold value, then corresponding territory element are deposited into abnormal area unit set A (gi) in until traversing entire trcheck；
Then compare A (gi) set territory element number and TR in the average area unit number of track judge whether track to be detected different
Often, and it may indicate that the sub-trajectory being abnormal；Finally decided whether to track carry out area according to the concrete condition of track to be detected
The track abnormality detection that domain divides again.
Although an embodiment of the present invention has been shown and described, it will be understood by those skilled in the art that：Not
In the case of being detached from the principle of the present invention and objective a variety of change, modification, replacement and modification can be carried out to these embodiments, this
The range of invention is limited by claim and its equivalent.
Claims (5)
1. a kind of abnormal track-detecting method based on region division, characterized in that include the following steps：
Step 1：Classify to the GPS historical trajectory datas of user, extracts the rail track feature of normal trace data, it will be normal
Track data carries out region division on geographical location, obtains track data region；
Step 2：The ready-portioned track data region of step 1 is subjected to track data territory element extension, with track data
The adjacent territory element in region is concluded into track data region, and historical track territory element sequence library is obtained；
Step 3：Region division is carried out to track data to be detected and obtains unit sequence tr in track regions to be measuredcheck={ g1,
g2,...,gn, wherein gnIndicate the territory element that track to be detected is passed through, n indicates territory element serial number, from historical track area
It is starting point that unit identical with the initiation region unit of track regions unit sequence to be measured is found in the unit sequence library of domain, and to be measured
The identical unit of termination area unit of track regions unit sequence is terminal, composition normal trace set TR={ tr1,tr2,
tr3,...,tri, wherein i is the number of track in TR | TR |, triIndicate i-th track, tri={ gi1,gi2,gi3,...,
gij, gijIndicate track triJ-th of the territory element passed through；
Step 4：Traverse unit sequence tr in track regions to be detectedcheckIn each unit lattice track, obtain each cell just
Supporting rate in normal practice trace set TR, and compared with threshold value, obtain abnormal area unit set A (gi)；The supporting rate
That is proportion function,|Inc(Tr,trcheck) | indicate track set Inc (TR, trcheck)
Tracking quantity, | Tr | indicate the quantity of normal trace set TR；Track abnormality detection is specially：
S41：Traverse trcheckEach compositing area unit, with normal trace set TR={ tr1,tr2,tr3,...,triIn
Track tri={ gi1,gi2,gi3,...,gijTerritory element compare, if track triWith trcheckRegion having the same
Unit, then by track triIt records and deposits into track set Inc (TR, trcheck)；
S42：For each territory element g of track to be detectedi, calculate Inc (TR, the tr of each territory elementcheck)
Proportion function of the tracking quantity in normal trace set TR,Judge this according to proportion function
Whether territory element is track trcheckIn normal trace point, as Sup (Tr, trcheck)<θ, θ are threshold value, 0≤θ≤1, by this
Territory element is put into abnormal area unit set A (gi) in, and work as Sup (Tr, trcheckThis territory element is put by) >=θ
Normal region unit set N (gi) in；
Step 5：According to normal trace set TR and abnormal area unit set A (gi) length relation judge track to be detected
Abnormal conditions.
2. a kind of abnormal track-detecting method based on region division according to claim 1, it is characterized in that：Further include to
The step of GPS historical trajectory datas at family and track data to be detected carry out subdivided region and abnormality detection.
3. a kind of abnormal track-detecting method based on region division according to claim 1 or claim 2, it is characterized in that：It is described to carry
Take the rail track feature of normal trace data include the longitude of track, track latitude and timestamp.
4. a kind of abnormal track-detecting method based on region division according to claim 1 or claim 2, it is characterized in that：It is described to sentence
Break track to be detected abnormal conditions method be：
Obtain abnormal area unit set A (gi) after, when the quantity of abnormal area unit | A (gi) | more than one preset value
When, i.e. abnormal area unit set A (gi) number of elements | A (gi) | when more, which is abnormal；Conversely, the track is just
Often.
5. a kind of abnormal track-detecting method based on region division according to claim 4, it is characterized in that：To the exception
Territory element set A (gi) carry out it is abnormal when judging, using calculating exceptional value R (trcheck) mode judged, exceptional valueWherein α is constants, and k is the territory element with each tracks normal trace set TR
The relevant constant of average value of number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610102351.2A CN105785411B (en) | 2016-02-24 | 2016-02-24 | A kind of abnormal track-detecting method based on region division |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610102351.2A CN105785411B (en) | 2016-02-24 | 2016-02-24 | A kind of abnormal track-detecting method based on region division |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105785411A CN105785411A (en) | 2016-07-20 |
| CN105785411B true CN105785411B (en) | 2018-10-12 |
Family
ID=56402913
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610102351.2A Active CN105785411B (en) | 2016-02-24 | 2016-02-24 | A kind of abnormal track-detecting method based on region division |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105785411B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108092840B (en) * | 2016-11-23 | 2021-08-31 | 中国移动通信集团广东有限公司 | Network rate segmentation processing method and device |
| CN106643734B (en) * | 2016-12-30 | 2019-01-11 | 中国科学院电子学研究所 | The hierarchical processing method of space-time trajectory data |
| WO2019196934A1 (en) * | 2018-04-13 | 2019-10-17 | Shanghai Truthvision Information Technology Co., Ltd. | System and method for abnormal scene detection |
| CN108804539B (en) * | 2018-05-08 | 2022-03-18 | 山西大学 | Track anomaly detection method under time and space double view angles |
| CN109034181A (en) * | 2018-06-05 | 2018-12-18 | 广州杰赛科技股份有限公司 | The classification method and device of motion track, equipment, storage medium |
| CN110634264A (en) * | 2018-06-22 | 2019-12-31 | 镇江雅迅软件有限责任公司 | Children prevent system of wandering away based on indoor positioning technique |
| CN109218985B (en) * | 2018-08-08 | 2020-09-22 | 上海中交水运设计研究有限公司 | Ship operation behavior abnormity detection method and system based on pattern similarity |
| CN109726737B (en) * | 2018-11-27 | 2020-11-10 | 武汉极意网络科技有限公司 | Track-based abnormal behavior detection method and device |
| CN111862586B (en) * | 2019-12-11 | 2021-10-15 | 北京嘀嘀无限科技发展有限公司 | Method and device for determining abnormal road section of road area and storage medium |
| CN111563137B (en) * | 2020-04-28 | 2022-05-17 | 厦门市美亚柏科信息股份有限公司 | Analysis method and system for coincident track |
| CN111474565A (en) * | 2020-05-20 | 2020-07-31 | 上海评驾科技有限公司 | Method for judging illegal plugging condition of road transport vehicle satellite positioning system terminal |
| CN111798971A (en) * | 2020-07-15 | 2020-10-20 | 智博云信息科技(广州)有限公司 | Medical information data processing method, system, electronic device and medium |
| CN111882873B (en) * | 2020-07-22 | 2022-01-28 | 平安国际智慧城市科技股份有限公司 | Track anomaly detection method, device, equipment and medium |
| CN114639216A (en) * | 2022-02-18 | 2022-06-17 | 国政通科技有限公司 | Specific personnel track area analysis early warning system and method |
| CN114822040B (en) * | 2022-06-23 | 2022-11-11 | 南京城建隧桥智慧管理有限公司 | Good neighbor set construction method for assisting mobile node position anomaly detection |
| CN115936561A (en) * | 2022-11-18 | 2023-04-07 | 广州云达供应链管理有限公司 | Logistics vehicle track operation abnormity monitoring method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8285060B2 (en) * | 2009-08-31 | 2012-10-09 | Behavioral Recognition Systems, Inc. | Detecting anomalous trajectories in a video surveillance system |
| CN104700434B (en) * | 2015-03-27 | 2017-10-31 | 北京交通大学 | A kind of crowd movement track method for detecting abnormality for labyrinth scene |
-
2016
- 2016-02-24 CN CN201610102351.2A patent/CN105785411B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN105785411A (en) | 2016-07-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105785411B (en) | A kind of abnormal track-detecting method based on region division | |
| CN110400332A (en) | A kind of target detection tracking method, device and computer equipment | |
| CN101827002B (en) | Concept drift detection method of data flow classification | |
| CN109684384B (en) | Trajectory data space-time density analysis system and analysis method thereof | |
| CN110188093A (en) | A kind of data digging system being directed to AIS information source based on big data platform | |
| CN107705324A (en) | A kind of video object detection method based on machine learning | |
| CN106600960A (en) | Traffic travel origin and destination identification method based on space-time clustering analysis algorithm | |
| CN104899263A (en) | Ship trajectory mining, analysis and monitoring method based on specific region | |
| Vajakas et al. | Trajectory reconstruction from mobile positioning data using cell-to-cell travel time information | |
| CN110426037A (en) | A kind of pedestrian movement track real time acquiring method under enclosed environment | |
| WO2020111934A1 (en) | A method and system for detection of natural disaster occurrence | |
| CN114998744B (en) | Agricultural machinery track field dividing method and device based on motion and vision dual-feature fusion | |
| CN105323024A (en) | Network signal intensity detecting and fusing method | |
| CN110674887A (en) | End-to-end road congestion detection algorithm based on video classification | |
| Rodrigues et al. | Impact of crowdsourced data quality on travel pattern estimation | |
| CN111475746B (en) | Point-of-interest mining method, device, computer equipment and storage medium | |
| CN111881233B (en) | Distributed point cloud map construction method and device, server and computer readable storage medium | |
| CN110503032A (en) | Individual important place detection method based on monitoring camera track data | |
| CN114417981A (en) | Intelligent river length patrol system | |
| CN113222040A (en) | Marine fixed target identification method and device, computer equipment and storage medium | |
| CN102945360A (en) | Method for detecting headwind zone of Doppler radar image based on morphological and logic operations | |
| CN114676907B (en) | Regional geological disaster early warning method and device, storage medium and equipment | |
| Gai et al. | Research on water quality spatiotemporal forecasting model based on ST-BIGRU-SVR neural network | |
| CN117517868A (en) | Lightning stroke fault matching method and system based on distributed measurement data | |
| CN113516850B (en) | Pipeline traffic flow data acquisition method based on space syntactic analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |