CN103996051A - Method for automatically detecting abnormal behaviors of video moving object based on change of movement features - Google Patents
Method for automatically detecting abnormal behaviors of video moving object based on change of movement features Download PDFInfo
- Publication number
- CN103996051A CN103996051A CN201410196646.1A CN201410196646A CN103996051A CN 103996051 A CN103996051 A CN 103996051A CN 201410196646 A CN201410196646 A CN 201410196646A CN 103996051 A CN103996051 A CN 103996051A
- Authority
- CN
- China
- Prior art keywords
- abnormal behaviour
- particle
- particles
- feature
- motion
- 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
Landscapes
- Image Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a method for automatically detecting abnormal behaviors of a video moving object based on change of movement features. In the method, the change state of the movement features of the video moving object can be reflected according to flow of particles in a Lagrangian particle dynamical system, extraction and clustering analysis are carried out on movement features of the particles to determine the intimacy degree of the classes of the movement features of the particles, and whether the abnormal behaviors of the video moving object occur is automatically detected based on the fact that when the abnormal behaviors of the video moving object occur, the classes of the movement features of the particles of the video moving object are different form the classes of movement features of particles with normal behaviors. By means of the method, it is not required that the moving object is tracked and abnormal behavior samples of the moving object are collected in advance and trained, and automatic detection of the abnormal behaviors of the video moving object can be achieved under various conditions.
Description
Technical field
The present invention relates to a kind of Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature, for public safety and strick precaution and video digital images analysis and understanding.Belong to intelligent information processing technology field.
Background technology
Day by day complicated along with the rapid growth of urban population and town environment, the cities and towns burst social security events such as Mass disturbance, riot, the attack of terrorism, are having a strong impact on cities and towns public safety.Construction is harmonious, safety is social, and oneself becomes an important topic of current international community.
In recent years, along with safety problem is subject to social growing interest, the demand of video monitoring system expands gradually, by the target real time monitoring to stop or process in special scenes, and the image information arriving according to the observation, the intelligent vision monitoring technology of Moving Objects behavior being carried out to semantic analysis and understanding has caused domestic extensive concern.The abnormal behaviour realizing in sequence of video images detects automatically, is still up to now one and has challenging work.On the one hand, there are a lot of differences in the definition of abnormal behaviour, is not yet also difficult to formulate unified specification.Wherein, some scholars will seldom occur or very short behavior of duration is referred to as abnormal behaviour, as falls or collision etc.; Some scholars is pre-defined a series of normal behaviours, are judged to be abnormal behaviour if find with the unmatched behavior of all normal behaviours; On the other hand, the abnormal behaviour in video sequence is of a great variety, and along with passage of time, abnormal behaviour also will change, and change normal behaviour into.At present, anomaly detection method is mainly divided into two classes: the one, based on the tracking of Moving Objects, by the movement locus of continuous detection Moving Objects, carry out abnormal behaviour detection.This class methods abnormal behaviour detection effect depends on the result of moving object tracking to a great extent, because actual scene is complicated and changeable and have blocking mutually and certainly blocking in various degree, cause effective moving object tracking difficulty, abnormal behaviour testing result is undesirable.The 2nd, based on learning training, by gathering in advance Moving Objects abnormal behaviour video sequence, set up abnormal behaviour Sample Storehouse, by unknown video sequence being detected and mating contrast, determine whether as abnormal behaviour.Because abnormal behaviour is of a great variety, and pass and constantly change in time, cause Moving Objects abnormal behaviour sample collection difficulty, abnormal behaviour testing result robustness is low.
Summary of the invention
The object of the invention is to obviously be affected by the external environment for current Moving Objects anomaly detection method result, the problems such as abnormal behaviour testing result robustness is low, provide a kind of Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature.It is can reflecting video Moving Objects according to flowing of particle in Lagrangian particle dynamic system motion feature variable condition, by Particles Moving feature extraction, Particles Moving feature is carried out to cluster analysis, determine other close and distant degree of Particles Moving feature class, during based on Moving Objects generation abnormal behaviour, the Particles Moving feature classification of its Particles Moving feature classification and normal behaviour there are differences, automatically detect Moving Objects in Video Sequences whether abnormal behaviour occurs, improve dirigibility and validity that Moving Objects in Video Sequences abnormal behaviour detects.
For achieving the above object, design of the present invention is: motion feature variable condition that can reflecting video Moving Objects according to flowing of particle in Lagrangian particle dynamic system, by Particles Moving feature extraction, Particles Moving feature is carried out to cluster analysis, determine other close and distant degree of Particles Moving feature class, during based on Moving Objects generation abnormal behaviour, the Moving Objects abnormal behaviour being built by particle flow field has different cluster modes in the feature space in particle flow field from its normal behaviour, automatically detect Moving Objects in Video Sequences whether abnormal behaviour occurs, realize Moving Objects in Video Sequences because of motion feature change abnormal behaviour automatically detect.
To achieve the above object, the present invention adopts following technical proposals:
A kind of abnormal behaviour automatic testing method changing based on Moving Objects in Video Sequences motion feature, it is characterized in that can reflecting video Moving Objects according to flowing of particle in Lagrangian particle dynamic system motion feature variable condition, by Particles Moving feature extraction, Particles Moving feature is carried out to cluster analysis, determine other close and distant degree of Particles Moving feature class, during based on Moving Objects generation abnormal behaviour, the Particles Moving feature classification of its Particles Moving feature classification and normal behaviour there are differences, automatically detect Moving Objects in Video Sequences whether abnormal behaviour occurs, concrete steps are as follows:
(1) start Moving Objects abnormal behaviour detected image acquisition system: gather video image;
(2) Particles Moving Flow Field Calculation;
(3) Particles Moving feature extraction;
(4) Particles Moving feature clustering;
(5) abnormal behaviour detects.
The concrete operation step of described step (2) Particles Moving Flow Field Calculation is as follows:
(2-1) optical flow computation: by camera acquisition
tmoment continuous adjacent two two field pictures, calculate optical flow field in level
xthe component of direction
u(
t):
with optical flow field vertical
ythe component of direction
v(
t):
;
(2-2) particle Flow Field Calculation:
Wherein,
,
represent respectively particle
p?
tbe engraved in level at+1 o'clock
xdirection is with vertical
yposition in direction,
,
represent respectively particle
p?
ttime be engraved in level
xdirection is with vertical
yposition in direction.
The concrete operation step of described step (3) Particles Moving feature extraction is as follows:
(3-1) average movement velocity calculates: exist according to the definite particle of step (2)
tmoment present position, calculates particle in a period of time
tinterior average movement velocity
:
;
(3-2) move distance calculates: according to the definite particle of step (2) in motion initial time and elapsed time
trear present position, calculates particle in the time
tthe distance of interior motion
d:
;
(3-3) direction of motion and direction histogram calculate: exist according to the definite particle of step (2)
tmoment present position, calculates
tthe direction of motion of moment particle
θ:
, and calculated direction histogram
h:
, wherein,
hfor histogram operational symbol.
The concrete operation step of described step (4) Particles Moving feature clustering is as follows:
(4-1) motion feature initial clustering: the Particles Moving feature definite according to step (3), comprising: average movement velocity
, move distance
d, direction of motion
qand direction histogram
h, constituent particle motion feature initial sets
f i :
f i =
,
d i , q i , h i ,
i=1,2 ...,
n, wherein,
nfor the product of image level size and vertical dimension;
(4-2) motion feature rough classification cluster: the Particles Moving feature initial sets definite according to step (4-1)
f i , from
f i middle searching is a pair of has similarity measurement value
dminimum subclass
? j with
? k ,
j=1,2 ..,
n;
k=1,2 ...,
n;
(4-3) motion feature subfractionation cluster: according to the definite subclass of step (4-2)
? j with
? k , will
? j be incorporated to
? k , and from
f i middle deletion subclass
? j , meanwhile,
jalso from index set
i:
i=1,2 ..., in n, delete.If
iradix equal set cluster numbers
ctime, stop calculating, otherwise, repeating step (4-2) and step (4-3).
The concrete operation step that described step (5) abnormal behaviour detects is as follows:
(5-1) abnormal behaviour is slightly judged: the Particles Moving feature initial sets definite according to step (4-1)
f i , when
f i in have while not belonging in step (4) any class in rough classification cluster and subfractionation cluster, be labeled as exception class subclass
? z (
z=1,2 ..,
n), and there is change in the behavior pattern of preliminary judgement Moving Objects;
(5-2) abnormal behaviour is determined: the exception class subclass in calculation procedure (5-1)
? z in number of particles
nif, number of particles
nbe greater than setting threshold
a, determine that abnormal behaviour has occurred Moving Objects.
Principle of the present invention is as follows:
In technical scheme of the present invention, motion feature is considered as to dynamic system non-periodic, it shows as a time dependent flow field, the motion feature providing according to dynamic particles in flow field is analyzed the motion state of Moving Objects, adopt and detect motion feature patterns of change, realize Moving Objects in Video Sequences abnormal behaviour and automatically detect.
Due to flowing of particle in Lagrangian particle dynamic system can reflecting video Moving Objects motion state especially the characteristic of fluid of Moving Objects abnormal behaviour change, therefore, utilize motion state and the feature thereof of Moving Objects in dynamic particles stream description video image, contribute to obtain efficient Moving Objects abnormal behavior.
If
for Lagrangian particle is at initial position
track,
for particle is through the time
after movement locus.
Wherein,
with
be respectively starting condition
under light stream in level
xdirection is moved with vertical
ythe estimation component of direction motion.
Consider following differential equation of first order:
y ¢=?
f(
t,?
y),
y(
t 0)?=?
y 0
Adopt Runge-Kutte method to solve the above-mentioned differential equation, thereby obtain particle respectively in level
xdirection is with vertical
yposition in direction:
Exist according to above-mentioned determined particle
tmoment present position
, calculate particle in a period of time
tinterior average movement velocity
:
, and particle is in the time
tinterior move distance
d:
, Particles Moving direction
q:
with particle direction histogram
h:
h=
h(
q).
By above-mentioned determined Particles Moving feature (
,
d,
q,
h) form
ndimension Particles Moving feature initial sets
f i :
f i =
,
d i , q i , h i ,
i=1,2 ...,
n, wherein,
nfor the product of image level size and vertical dimension.
According to above-mentioned definite initial sets
f i carry out hierarchical clustering, when initial, arrange
? l =
f i , "
l?
i,
i=1,2 ...,
n.
Set
? l |
l?
ithe a pair of cluster subclass that meets following condition of middle searching
? j with
? k .
Wherein,
be
? j with
? k between similarity measurement.
Will
? j be incorporated to
? k , and from set
? l middle deletion
? j , meanwhile,
jalso from index set
imiddle deletion.If
iradix equal set cluster numbers
ctime, stop calculating, otherwise repeat above-mentioned similarity measurement computation process.
When Particles Moving feature initial sets
f i while not belonging to any class in above-mentioned hierarchical clustering, be labeled as exception class subclass
? z (
z=1,2 ..,
n), and by exception class subclass
? z in number of particles
nbe greater than setting threshold
atime, judge that abnormal behaviour has occurred Moving Objects.
The present invention compared with prior art, there is following apparent outstanding substantive distinguishing features and remarkable advantage: the present invention is by introducing Lagrangian particle dynamic system, utilize the effectively motion state of reflecting video Moving Objects that flows of particle in dynamic system, thereby can extract the motion feature of stable Moving Objects in Video Sequences, according to Moving Objects in the variation of feature space motor pattern of doing more physical exercises, automatically detect Moving Objects in Video Sequences and whether have abnormal behaviour, solve existing method in the time of Moving Objects in Video Sequences abnormal behaviour, be limited to specified conditions or environment, and to dynamic scene sensitive, noise is large, the deficiency that abnormal behaviour testing result is undesirable, improve the robustness that Moving Objects in Video Sequences abnormal behaviour detects, can adapt to the automatic detection of Moving Objects in Video Sequences abnormal behaviour under various complicated condition.Method of the present invention is easy, flexible, easily realization.
Brief description of the drawings
Fig. 1 is the flowsheet of the inventive method.
Fig. 2 is an original two field picture of one embodiment of the invention.
Fig. 3 is the optical flow field image calculating from original video sequence consecutive frame of Fig. 2 example.
Fig. 4 is the corresponding frame particle flux image of Fig. 2 example.
Fig. 5 is a frame abnormal behaviour testing result image of one embodiment of the invention.
Embodiment
Details are as follows by reference to the accompanying drawings for the preferred embodiments of the present invention:
Embodiment mono-:
Referring to Fig. 1, the Moving Objects in Video Sequences abnormal behaviour automatic testing method that this changes based on motion feature, is characterized in that operation steps is as follows:
(1) start Moving Objects abnormal behaviour detected image acquisition system: gather video image;
(2) Particles Moving Flow Field Calculation;
(3) Particles Moving feature extraction;
(4) Particles Moving feature clustering;
(5) abnormal behaviour detects.
Embodiment bis-:
The present embodiment and embodiment mono-are basic identical, and special feature is as follows:
The concrete operation step of described step (2) Particles Moving Flow Field Calculation is as follows:
(2-1) optical flow computation: by camera acquisition
tmoment continuous adjacent two two field pictures, calculate optical flow field in level
xthe component of direction
u(
t):
with optical flow field vertical
ythe component of direction
v(
t):
;
(2-2) particle Flow Field Calculation:
Wherein,
,
represent respectively particle
p?
tbe engraved in level at+1 o'clock
xdirection is with vertical
yposition in direction,
,
represent respectively particle
p?
ttime be engraved in level
xdirection is with vertical
yposition in direction.
The concrete operation step of described step (3) Particles Moving feature extraction is as follows:
(3-1) average movement velocity calculates: exist according to the definite particle of step (2)
tmoment present position, calculates particle in a period of time
tinterior average movement velocity
:
;
(3-2) move distance calculates: according to the definite particle of step (2) in motion initial time and elapsed time
trear present position, calculates particle in the time
tthe distance of interior motion
d:
;
(3-3) direction of motion and direction histogram calculate: exist according to the definite particle of step (2)
tmoment present position, calculates
tthe direction of motion of moment particle
θ:
, and calculated direction histogram
h:
, wherein,
hfor histogram operational symbol.
The concrete operation step of described step (4) Particles Moving feature clustering is as follows:
(4-1) motion feature initial clustering: the Particles Moving feature definite according to step (3), comprising: average movement velocity
, move distance
d, direction of motion
qand direction histogram
h, constituent particle motion feature initial sets
f i :
f i =
,
d i , q i , h i ,
i=1,2 ...,
n, wherein,
nfor the product of image level size and vertical dimension;
(4-2) motion feature rough classification cluster: the Particles Moving feature initial sets definite according to step (4-1)
f i , from
f i middle searching is a pair of has similarity measurement value
dminimum subclass
? j with
? k ,
j=1,2 ..,
n;
k=1,2 ...,
n;
(4-3) motion feature subfractionation cluster: according to the definite subclass of step (4-2)
? j with
? k , will
? j be incorporated to
? k , and from
f i middle deletion subclass
? j , meanwhile,
jalso from index set
i:
i=1,2 ..., in n, delete.If
iradix equal set cluster numbers
ctime, stop calculating, otherwise, repeating step (4-2) and step (4-3).
The concrete operation step that described step (5) abnormal behaviour detects is as follows:
(5-1) abnormal behaviour is slightly judged: the Particles Moving feature initial sets definite according to step (4-1)
f i , when
f i in have while not belonging in step (4) any class in rough classification cluster and subfractionation cluster, be labeled as exception class subclass
? z (
z=1,2 ..,
n), and there is change in the behavior pattern of preliminary judgement Moving Objects;
(5-2) abnormal behaviour is determined: the exception class subclass in calculation procedure (5-1)
? z in number of particles
nif, number of particles
nbe greater than setting threshold
a, determine that abnormal behaviour has occurred Moving Objects.
Embodiment tri-:
Referring to Fig. 1~Fig. 5, the present embodiment is: running program as shown in Figure 1, an original two field picture of this example as shown in Figure 2, sequence of video images shown in Fig. 2 is adopted to the motion state of Moving Objects in dynamic particles stream description video image, set up the motion feature space based on particle flow field, have different Clusterings according to Moving Objects abnormal behaviour from the feature space of its normal behaviour in particle flow field, the abnormal behaviour that realizes the variation of Moving Objects in Video Sequences motion feature detects automatically.Concrete steps are as follows:
(1) start Moving Objects abnormal behaviour detected image acquisition system: gather video image;
(2) Particles Moving Flow Field Calculation: by continuous adjacent two two field pictures of camera acquisition (image size: 320 ' 240), calculate optical flow field (
u(
t),
v(
t)), as shown in Figure 3; To the image shown in Fig. 3, calculate particle flow field, as shown in Figure 4;
(3) Particles Moving feature extraction: in each moment present position, calculate the average movement velocity of particle in a period of times 25 frame according to the determined particle of step (2)
:
, move distance
d:
, direction of motion
θ:
, and direction histogram
h:
;
(4) Particles Moving feature clustering: according to the determined Particles Moving feature of step (3), comprising: average movement velocity
, move distance
d, direction of motion
qand direction histogram
h, form 76800 dimension Particles Moving feature initial sets
f i :
f i =
,
d i , q i , h i , i=1,2 ..., 76800, adopt Euclidean distance to carry out hierarchical clustering as similarity measurement, and until cluster numbers is 2 o'clock, stop calculating, otherwise, repeat the above-mentioned hierarchical clustering that is similarity measurement based on Euclidean distance;
Abnormal behaviour detects: according to Particles Moving feature initial sets
f i :
f i =
,
d i , q i , h i , i=1,2 ..., 76800, when
f i in while thering is any hierarchical clustering not belonging in step (4), be labeled as exception class subclass
? z , and there is change in the behavior pattern of preliminary judgement Moving Objects; And and then calculating exception class subclass
? z in number of particles be 1650, its number of particles 1650 is greater than setting threshold 1000, shows that abnormal behaviour has occurred Moving Objects, as Fig. 5.
Claims (5)
1. the Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature, is characterized in that operation steps is as follows:
Start Moving Objects abnormal behaviour detected image acquisition system: gather video image;
Particles Moving Flow Field Calculation;
Particles Moving feature extraction;
Particles Moving feature clustering;
Abnormal behaviour detects.
2. the Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature according to claim 1, is characterized in that: the concrete operation step of described step (2) Particles Moving Flow Field Calculation is as follows:
Optical flow computation: by camera acquisition
tmoment continuous adjacent two two field pictures, calculate optical flow field in level
xthe component of direction
u(
t):
with optical flow field vertical
ythe component of direction
v(
t):
;
Particle Flow Field Calculation:
Wherein,
,
represent respectively particle
p?
tbe engraved in level at+1 o'clock
xdirection is with vertical
yposition in direction,
,
represent respectively particle
p?
ttime be engraved in level
xdirection is with vertical
yposition in direction.
3. the Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature according to claim 1, is characterized in that: the concrete operation step of described step (3) Particles Moving feature extraction is as follows:
Average movement velocity calculates: exist according to the definite particle of step (2)
tmoment present position, calculates particle in a period of time
tinterior average movement velocity
:
;
Move distance calculates: according to the definite particle of step (2) in motion initial time and elapsed time
trear present position, calculates particle in the time
tthe distance of interior motion
d:
;
Direction of motion and direction histogram calculate: exist according to the definite particle of step (2)
tmoment present position, calculates
tthe direction of motion of moment particle
θ:
, and calculated direction histogram
h:
, wherein,
hfor histogram operational symbol.
4. the Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature according to claim 1, is characterized in that: the concrete operation step of described step (4) Particles Moving feature clustering is as follows:
Motion feature initial clustering: the Particles Moving feature definite according to step (3), comprising: average movement velocity
, move distance
d, direction of motion
qand direction histogram
h, constituent particle motion feature initial sets
f i :
f i =
,
d i , q i , h i ,
i=1,2 ...,
n, wherein,
nfor the product of image level size and vertical dimension;
Motion feature rough classification cluster: the Particles Moving feature initial sets definite according to step (4-1)
f i , from
f i middle searching is a pair of has similarity measurement value
dminimum subclass
? j with
? k ,
j=1,2 ..,
n;
k=1,2 ...,
n;
Motion feature subfractionation cluster: according to the definite subclass of step (4-2)
? j with
? k , will
? j be incorporated to
? k , and from
f i middle deletion subclass
? j , meanwhile,
jalso from index set
i:
i=1,2 ...,
nmiddle deletion; If
iradix equal set cluster numbers
ctime, stop calculating, otherwise, repeating step (4-2) and step (4-3).
5. the Moving Objects in Video Sequences abnormal behaviour automatic testing method changing based on motion feature according to claim 1, is characterized in that: the concrete operation step that described step (5) abnormal behaviour detects is as follows:
Abnormal behaviour is slightly judged: the Particles Moving feature initial sets definite according to step (4-1)
f i , when
f i in have while not belonging in step (4) any class in rough classification cluster and subfractionation cluster, be labeled as exception class subclass
? z (
z=1,2 ..,
n), and there is change in the behavior pattern of preliminary judgement Moving Objects;
Abnormal behaviour is determined: the exception class subclass in calculation procedure (5-1)
? z in number of particles
nif, number of particles
nbe greater than setting threshold
a, determine that abnormal behaviour has occurred Moving Objects.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410196646.1A CN103996051A (en) | 2014-05-12 | 2014-05-12 | Method for automatically detecting abnormal behaviors of video moving object based on change of movement features |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410196646.1A CN103996051A (en) | 2014-05-12 | 2014-05-12 | Method for automatically detecting abnormal behaviors of video moving object based on change of movement features |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103996051A true CN103996051A (en) | 2014-08-20 |
Family
ID=51310212
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410196646.1A Pending CN103996051A (en) | 2014-05-12 | 2014-05-12 | Method for automatically detecting abnormal behaviors of video moving object based on change of movement features |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103996051A (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105469054A (en) * | 2015-11-25 | 2016-04-06 | 天津光电高斯通信工程技术股份有限公司 | Model construction method of normal behaviors and detection method of abnormal behaviors |
CN106157325A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Group abnormality behavioral value method and system |
CN106157326A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Group abnormality behavioral value method and system |
CN106156706A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Pedestrian's anomaly detection method |
CN106768837A (en) * | 2016-12-08 | 2017-05-31 | 国家海洋局北海预报中心 | Persons falling in water sea drift orbit Forecasting Methodology |
CN107292266A (en) * | 2017-06-21 | 2017-10-24 | 吉林大学 | A kind of vehicle-mounted pedestrian area estimation method clustered based on light stream |
CN108052859A (en) * | 2017-10-31 | 2018-05-18 | 深圳大学 | A kind of anomaly detection method, system and device based on cluster Optical-flow Feature |
CN108230305A (en) * | 2017-12-27 | 2018-06-29 | 浙江新再灵科技股份有限公司 | Method based on the detection of video analysis staircase abnormal operating condition |
CN110378200A (en) * | 2019-06-03 | 2019-10-25 | 特斯联(北京)科技有限公司 | A kind of intelligent security guard prompt apparatus and method for of Behavior-based control feature clustering |
CN110536114A (en) * | 2019-08-05 | 2019-12-03 | 特斯联(北京)科技有限公司 | A kind of the parking lot CCTV monitoring system and method for Intelligent target tracking |
CN111310583A (en) * | 2020-01-19 | 2020-06-19 | 中国科学院重庆绿色智能技术研究院 | Vehicle abnormal behavior identification method based on improved long-term and short-term memory network |
US10783765B2 (en) | 2018-09-13 | 2020-09-22 | Wistron Corporation | Falling detection method and electronic system using the same |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050105764A1 (en) * | 2003-11-17 | 2005-05-19 | Mei Han | Video surveillance system with connection probability computation that is a function of object size |
CN101751679A (en) * | 2009-12-24 | 2010-06-23 | 北京中星微电子有限公司 | Sorting method, detecting method and device of moving object |
-
2014
- 2014-05-12 CN CN201410196646.1A patent/CN103996051A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050105764A1 (en) * | 2003-11-17 | 2005-05-19 | Mei Han | Video surveillance system with connection probability computation that is a function of object size |
CN101751679A (en) * | 2009-12-24 | 2010-06-23 | 北京中星微电子有限公司 | Sorting method, detecting method and device of moving object |
Non-Patent Citations (2)
Title |
---|
王乔: "运动人体的异常行为检测研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
石华伟: "人体运动的跟踪与识别", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106157325A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Group abnormality behavioral value method and system |
CN106157326A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Group abnormality behavioral value method and system |
CN106156706A (en) * | 2015-04-07 | 2016-11-23 | 中国科学院深圳先进技术研究院 | Pedestrian's anomaly detection method |
CN106157326B (en) * | 2015-04-07 | 2020-10-27 | 中国科学院深圳先进技术研究院 | Group abnormal behavior detection method and system |
CN106156706B (en) * | 2015-04-07 | 2020-05-19 | 中国科学院深圳先进技术研究院 | Pedestrian abnormal behavior detection method |
CN105469054B (en) * | 2015-11-25 | 2019-05-07 | 天津光电高斯通信工程技术股份有限公司 | The model building method of normal behaviour and the detection method of abnormal behaviour |
CN105469054A (en) * | 2015-11-25 | 2016-04-06 | 天津光电高斯通信工程技术股份有限公司 | Model construction method of normal behaviors and detection method of abnormal behaviors |
CN106768837A (en) * | 2016-12-08 | 2017-05-31 | 国家海洋局北海预报中心 | Persons falling in water sea drift orbit Forecasting Methodology |
CN107292266A (en) * | 2017-06-21 | 2017-10-24 | 吉林大学 | A kind of vehicle-mounted pedestrian area estimation method clustered based on light stream |
CN108052859A (en) * | 2017-10-31 | 2018-05-18 | 深圳大学 | A kind of anomaly detection method, system and device based on cluster Optical-flow Feature |
CN108230305A (en) * | 2017-12-27 | 2018-06-29 | 浙江新再灵科技股份有限公司 | Method based on the detection of video analysis staircase abnormal operating condition |
US10783765B2 (en) | 2018-09-13 | 2020-09-22 | Wistron Corporation | Falling detection method and electronic system using the same |
CN110378200A (en) * | 2019-06-03 | 2019-10-25 | 特斯联(北京)科技有限公司 | A kind of intelligent security guard prompt apparatus and method for of Behavior-based control feature clustering |
CN110536114A (en) * | 2019-08-05 | 2019-12-03 | 特斯联(北京)科技有限公司 | A kind of the parking lot CCTV monitoring system and method for Intelligent target tracking |
CN111310583A (en) * | 2020-01-19 | 2020-06-19 | 中国科学院重庆绿色智能技术研究院 | Vehicle abnormal behavior identification method based on improved long-term and short-term memory network |
CN111310583B (en) * | 2020-01-19 | 2023-02-10 | 中国科学院重庆绿色智能技术研究院 | Vehicle abnormal behavior identification method based on improved long-term and short-term memory network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103996051A (en) | Method for automatically detecting abnormal behaviors of video moving object based on change of movement features | |
CN103530874B (en) | People stream counting method based on Kinect | |
CN108074244B (en) | Safe city traffic flow statistical method integrating deep learning and background difference method | |
CN106469313B (en) | A kind of detection method of small target of caliber adaptive space-time domain filtering | |
CN103473533B (en) | Moving Objects in Video Sequences abnormal behaviour automatic testing method | |
CN107067413B (en) | A kind of moving target detecting method of time-space domain statistical match local feature | |
CN109559302A (en) | Pipe video defect inspection method based on convolutional neural networks | |
CN104978567B (en) | Vehicle checking method based on scene classification | |
CN103942533A (en) | Urban traffic illegal behavior detection method based on video monitoring system | |
CN107506734A (en) | One kind of groups unexpected abnormality event detection and localization method | |
CN102065275B (en) | Multi-target tracking method in intelligent video monitoring system | |
CN103678299A (en) | Method and device for monitoring video abstract | |
CN108537829B (en) | Monitoring video personnel state identification method | |
CN108038866A (en) | A kind of moving target detecting method based on Vibe and disparity map Background difference | |
CN111339917B (en) | Method for detecting glass in real scene | |
Karpagavalli et al. | Estimating the density of the people and counting the number of people in a crowd environment for human safety | |
CN108537157A (en) | A kind of video scene judgment method and device based on artificial intelligence classification realization | |
CN106991418A (en) | Winged insect detection method, device and terminal | |
CN107273815A (en) | A kind of individual behavior recognition methods and system | |
CN106372590A (en) | Sea surface ship intelligent tracking system and method based on machine vision | |
CN108986143A (en) | Target detection tracking method in a kind of video | |
CN101877135B (en) | Moving target detecting method based on background reconstruction | |
CN115188081B (en) | Complex scene-oriented detection and tracking integrated method | |
CN104200202A (en) | Upper human body detection method based on accumulative perceptron | |
Bhaskar et al. | Enhanced and effective parallel optical flow method for vehicle detection and tracking |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140820 |
|
WD01 | Invention patent application deemed withdrawn after publication |