CN107705326A - A kind of intrusion detection method that crosses the border in security sensitive region - Google Patents
A kind of intrusion detection method that crosses the border in security sensitive region Download PDFInfo
- Publication number
- CN107705326A CN107705326A CN201710833558.1A CN201710833558A CN107705326A CN 107705326 A CN107705326 A CN 107705326A CN 201710833558 A CN201710833558 A CN 201710833558A CN 107705326 A CN107705326 A CN 107705326A
- Authority
- CN
- China
- Prior art keywords
- mrow
- sensitive region
- security sensitive
- border
- background
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/254—Analysis of motion involving subtraction of images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The invention discloses a kind of intrusion detection method that crosses the border in security sensitive region, belong to Video security monitoring field, this method comprises the following steps:1) camera real-time image acquisition, and security sensitive region is set to the pickup area of camera;2) moving target information is extracted from the video flowing collected by background subtraction method;3) foreground image extracted, i.e. moving target are handled using morphologic filtering method;4) connected region profile in foreground image is extracted, and screens moving target, area, size or the undesirable set of connected domain of ratio are removed;5) judge whether moving target enters security sensitive region set in advance using vector cross product method.The inventive method can effectively filter out the intrusion event that crosses the border in monitor area, mitigate the work load of monitoring personnel, avoid the anomalous event caused by monitoring personnel work fatigue from failing to report.
Description
Technical field
The present invention relates to Video security monitoring field, more particularly to a kind of intrusion detection side of crossing the border in security sensitive region
Method.
Background technology
With the increase of the current size of population and the continuous expansion of city size, what modern society was faced happens suddenly, is abnormal
Event increases year by year so that the difficulty of security monitoring is also more and more prominent with importance.Especially for some security sensitive fields
Close, such as military base, prison, supply station, hospital, bank and market region, for the consideration of safety factor, administrative staff
Whether have to be understood that has abnormal personage or vehicle to enter the place.Traditional Video security monitoring is all taking human as master, by security
Personnel are directly viewable monitored picture.Developing rapidly and popularizing with Video Supervision Technique, camera quantity is more and more, in addition
The fatigability of human eye, it may lead to not reliably filter out anomalous event and be handled in time.Therefore, had using one kind
The intrusion detection method that crosses the border of effect is particularly important.
The content of the invention
For above-mentioned deficiency, the present invention proposes a kind of intrusion detection method that crosses the border in security sensitive region, passes through the back of the body first
Scape relief method detects the moving target in video, morphologic filtering and connected domain analysis is then carried out again, finally according to vector
Cross product method judges whether moving target enters security sensitive region set in advance, and this method can make monitoring system automatic screening
The intrusion event that crosses the border gone out in monitor area, mitigate the work load of monitoring personnel.
The technical solution adopted in the present invention is as follows:A kind of intrusion detection method that crosses the border in security sensitive region, including with
Lower step:
1) camera real-time image acquisition, and security sensitive region is set to the pickup area of camera;
2) moving target information is extracted from the video flowing collected by background subtraction method;
3) foreground image extracted, i.e. moving target are handled using morphologic filtering method;
4) connected region profile in foreground image is extracted, and screens moving target, area, size or ratio are not met will
The connected domain asked removes;
5) judge whether the moving target in foreground image enters security sensitive region, that is, judge a point whether at one
In quadrangle, judge whether angle has exceeded 180 degree using the directionality of two vector cross products, ifThen
Represent that two vectorial angles are less than 180 degree, i.e. point E existsClockwise direction;IfThen represent two
The angle of vector is more than 180 degree, i.e. point E existsCounter clockwise direction;It can thus be concluded that if
Then represent point E in vectorWithBetween;Therefore, Rule of judgment is passed through Whether set up, can obtain whether target point E comes into four surrounded by point A, B, C, D
Among the shape of side, the quadrangle that point A, B, C, D are surrounded is security sensitive region.
Further, the mode in setting security sensitive region has following several:1) the parameter configuration text of last time preservation is read
Part;2) delimited by user by mouse;3) set by user's input coordinate value.
Further, the background subtraction method is selected from mixed Gauss model method, VIBE algorithms, SACON algorithms, PBAS algorithms
In one kind.
Further, moving target information is extracted from video flowing using PBAS algorithms, detailed process includes:
2.1) background modeling
Initial value of the pixel value of N two field pictures as background model before collection:
B(xi)={ B1(xi),…,Bk(xi),…,BN(xi)}
Wherein, B (xi) represent pixel xiIn the background model at current time, Bk(xi) represent pixel xiThe picture at history k moment
Element value;
2.2) foreground detection
For any one pixel x in current image framei, judge that formula is:
Wherein, F (xi)=1 represents xiIt is judged as foreground point, F (xi)=0 represents xiIt is judged as background dot, num (.) table
Show the number for meeting condition in bracket, dist (.) represent the Euclidean distance between 2 points in bracket;
2.3) context update
Background model is updated with the pixel for being identified as background, context update is divided into following three steps:
2.3.1) renewal background model:From its background model B (xi) one pixel value of middle random selection, with I (xi) replace,
It is p (x that it, which replaces probability,i)=1/T (xi), T (xi) represent context update rate;
2.3.2 domain background model) is updated:From xiField in randomly choose pixel yi, with its current pixel value V
(yi) replace its background model B (yi) one pixel value of middle random selection, it is p (x that it, which replaces probability,i)=1/T (xi);
2.3.3 judgment threshold R (x) are adaptively adjustedi) and turnover rate T (xi):Calculate pixel x in N two field pictures in the pastiWith it
The average value of corresponding background sample minimum range represents background complexity, and to adjust prospect according to background complicacy self-adaptive judge
Threshold value R (xi) and background model turnover rate T (xi)。
Further, the foreground image come out using morphologic filtering method processing detection, it is specific as follows:By opening
Reflation is first corroded in computing, removes the profile of small noise and smooth motion object;First expanded by closed operation and corroded again, filled out
Fill cavity tiny inside foreground target.
The beneficial effects of the invention are as follows:
1. can be with work free of discontinuities in 24 hours, monitoring personnel only needs to carry out unusual condition processing when alarm
, human resources have been effectively saved, while avoid the omission of the anomalous event caused by monitoring personnel work fatigue;
2. moving target information is extracted from video flowing using adaptivenon-uniform sampling (PBAS) algorithm based on pixel so that
The inventive method can adapt to complex background environment, and robustness is stronger, and accuracy of detection is higher;
3. by morphologic filtering and connected domain analysis, influence of the noise to model is further reduced so that the present invention
Method can be applied to various scenes complicated and changeable.
Brief description of the drawings
Fig. 1 is the flow chart of the intrusion detection method of the invention that crosses the border;
Fig. 2 is to judge whether a point enters the schematic diagram of quadrangle using cross product method;
Fig. 3 is that one kind is crossed the border intrusion detection system structure block diagram;
Fig. 4 is the moving object detection result figure of the present invention;
Fig. 5 is the intrusion detection result figure of crossing the border of the present invention.
Embodiment
Below in conjunction with drawings and examples, the present invention is further illustrated.
As shown in figure 1, the present invention proposes a kind of intrusion detection method that crosses the border in security sensitive region, mainly including following step
Suddenly:
1) video is read in
Video data can be from camera collection realtime graphic or be stored in local data file,
The pretreatment such as medium filtering can be carried out to video data according to actual conditions.
2) security sensitive region is set
The setting in security sensitive region is carried out to the pickup area of camera, providing the user with 3 kinds of modes, to set safety quick
Sensillary area domain:1) parameter configuration files of last time preservation are read;2) delimited by user by mouse;3) by user's input coordinate value
Setting.
3) moving object detection
The present invention detects moving target using background subtraction method.At present, conventional background subtraction method has mixed Gaussian mould
Type method, VIBE algorithms, SACON algorithms, PBAS algorithms etc..In view of outdoor monitoring easily by illumination variation, sleety weather,
The influence of the uncertain factors such as float, according to the comparative analysis and test to various algorithms, the present invention uses PBAS algorithms
To extract moving target information from video flowing, the algorithm combines the advantage of two kinds of algorithms of SACON and VIBE, carries on the back by introducing
The measure of scape complexity, prospect judgment threshold and background model turnover rate are adaptively adjusted according to complex degree of background,
Detailed process includes:
3.1) background modeling
Initial value of the pixel value of N two field pictures as background model before collection:
B(xi)={ B1(xi),…,Bk(xi),…,BN(xi)}
Wherein, B (xi) represent pixel xiIn the background model at current time, Bk(xi) represent pixel xiThe picture at history k moment
Element value.
3.2) foreground detection
For any one pixel x in current image frameiIf its pixel value I (xi) with background model in N number of pixel
The Euclidean distance of value is less than judgment threshold R (xi) number be less than #min, then the point is judged to foreground point, is otherwise judged to carry on the back
Sight spot.
Judge that formula is:
Wherein, F (xi)=1 represents xiIt is judged as foreground point, F (xi)=0 represents xiIt is judged as background dot, num (.) table
Show the number for meeting condition in bracket, dist (.) represent the Euclidean distance between 2 points in bracket.
3.3) context update
Using the strategy of selective updating, i.e., being only identified as the pixel of background could be used for updating background model.
If pixel xiBackground dot is judged as, context update can be divided into following three steps:
3.3.1) renewal background model.From its background model B (xi) one pixel value of middle random selection, with I (xi) replace,
It is p (x that it, which replaces probability,i)=1/T (xi), T (xi) represent context update rate.
3.3.2 domain background model) is updated.From xiField in randomly choose pixel yi, with its current pixel value V
(yi) replace its background model B (yi) one pixel value of middle random selection, it is p (x that it, which replaces probability,i)=1/T (xi)。
3.3.3 judgment threshold R (x) are adaptively adjustedi) and turnover rate T (xi).Calculate pixel x in N two field pictures in the pastiWith it
The average value of corresponding background sample minimum range represents background complexity, and to adjust prospect according to background complicacy self-adaptive judge
Threshold value R (xi) and background model turnover rate T (xi)。
4) morphologic filtering
The foreground image (i.e. moving target) come out using morphologic filtering method processing detection:It is first rotten by opening operation
Reflation is lost, removes the profile of small noise and smooth motion object;First expanded by closed operation and corroded again, fill foreground target
Internal tiny cavity.
5) connected domain analysis
Connected region profile in foreground image is extracted, and the sieve such as contour area size, length-width ratio is set according to prior information
Moving target is selected, area, size or the undesirable set of connected domain of ratio are removed, done so as to eliminate noise to a certain extent
Disturb, improve Detection accuracy.
6) cross the border detection
Detection of crossing the border mainly judges whether foreground target enters security sensitive region, and its core is exactly to judge one
Whether point is in a quadrangle.Judge whether angle has exceeded 180 degree, such as Fig. 2 using the directionality of two vector cross products
It is shown, ifThen represent that two vectorial angles are less than 180 degree, i.e. point E existsClockwise direction;IfThen represent that two vectorial angles are more than 180 degree, i.e. point E existsCounter clockwise direction.It can thus be concluded that ifThen represent point E in vectorWithBetween.Therefore, Rule of judgment is passed throughWhether set up, target point E can be obtained whether
Through entering among the quadrangle surrounded by point A, B, C, D, the quadrangle that its midpoint A, B, C, D are surrounded is quick for the safety of user's setting
Sensillary area domain.
7) warning reminding
If detecting the intrusion event that crosses the border, picture is preserved, and warning message and picture are sent to user.
For above detection method, the present invention have also been devised one kind and cross the border intruding detection system, invasion of crossing the border of the invention
Detection method is realized in this crosses the border intruding detection system.
As shown in figure 3, the intruding detection system of crossing the border that the present invention designs includes:N camera, n embeded processor,
One server and m client, wherein n, m are positive integer, and a camera passes through with a corresponding embeded processor
Data wire is connected, and all embeded processors are connected by wired or wireless with server, and all clients are by wired
Or wirelessly it is connected with server;The intrusion detection method that crosses the border of the present invention can be transported on the embeded processor of front monitoring front-end
OK, can also run on the client.
The image in the real-time acquisition monitoring region of camera, and these view data are passed into front monitoring front-end and are attached thereto
Embeded processor;
Embeded processor receives the view data that camera sends over, and certain place is then carried out to view data
Reason, and upload onto the server;
Server receives the view data that each embeded processor sends over, and carries out classification to these data and deposit
Storage, it is easy to monitoring personnel to play back or collect evidence;
The information that client the reception server sends over, then monitoring personnel judge whether to occur by checking image
Cross the border intrusion event;Meanwhile monitoring personnel can also send control by user end to server and each embeded processor and refer to
Order, for changing, updating system configuration and parameter.
According to different network environment and work requirements, the intruding detection system of crossing the border can be set to two kinds of Working moulds
Formula:1. real-time monitoring mode;2. non real-time monitoring mode.
In real-time monitoring mode, embeded processor is only responsible for the view data of camera collection uploading to service
Device;Server is stored to image classification, and view data is transmitted into client;Client real-time display server sends over
View data.In this mode of operation, the intrusion detection method that crosses the border of the invention is run in the client, can partly be replaced
For the work of monitoring personnel, the automatic detection for the intrusion event that crosses the border is realized, has been effectively saved human resources, avoided due to prison
Anomalous event caused by control person works' fatigue is failed to report.
In non real-time monitoring mode, embeded processor enters the view data gathered using the inventive method to camera
Capable intrusion detection of crossing the border, if having detected target invasion, screen shot and testing result are preserved, and upload onto the server;
Then, server is to client push warning message and picture.In this mode of operation, client does not need real-time reception
From the view data of camera, only related warning message need to be received when FEP detects target invasion, this
The redundancy in monitor video is greatly reduced, reduces requirement of the video monitoring to data traffic.
The system can automatically read when starting and be loaded into the configuration parameter information of last time preservation;Can be certainly when stop
Dynamic preservation system relevant configured parameter information.
Embodiment:
Exemplified by having detected whether target illegal crossings greenbelt.As shown in figure 4, the moving target detecting method of the present invention
Accurate three moving targets that detected in monitoring scene.As shown in figure 5, the moving target detected is respectively labeled as
" 1 ", " 2 ", " 3 ", middle quadrangle greenbelt is set to security sensitive region, marked in video with dotted line.Due to No. 1 motion
Whether target judges No. 1 moving target bottom right angle point in security sensitive region upper left, by the vector cross product method of the present invention
Judge whether the target has crossed into the security sensitive region inside quadrangle.From fig. 5, it can be seen that when No. 1 fortune
When moving-target enters security sensitive region set in advance, system can to monitoring personnel send " Warning " warning message and
Screen shot, monitoring personnel are examined and handled to anomalous event further according to sectional drawing after receiving warning message, so can be with
Avoid monitoring personnel from dig-inning screen always, mitigate the work load of monitoring personnel significantly.
Claims (5)
1. a kind of intrusion detection method that crosses the border in security sensitive region, it is characterised in that comprise the following steps:
1) camera real-time image acquisition, and security sensitive region is set to the pickup area of camera;
2) moving target information is extracted from the video flowing collected by background subtraction method;
3) foreground image extracted, i.e. moving target are handled using morphologic filtering method;
4) connected region profile in foreground image is extracted, and screens moving target, by area, size or the undesirable set of company of ratio
Logical domain removes.
5) judge whether the moving target in foreground image enters security sensitive region, that is, judge a point whether in a quadrangle
It is interior, judge whether angle has exceeded 180 degree using the directionality of two vector cross products, ifThen represent two to
The angle of amount is less than 180 degree, i.e. point E existsClockwise direction;IfThen represent that two vectorial angles are more than
180 degree, i.e. point E existCounter clockwise direction;It can thus be concluded that ifThen represent point E in vectorWithBetween;Therefore, Rule of judgment is passed through
Whether set up, can obtain whether target point E is come among the quadrangle surrounded by point A, B, C, D, point A, B, C, D are surrounded
Quadrangle be security sensitive region.
A kind of 2. intrusion detection method that crosses the border in security sensitive region according to claim 1, it is characterised in that setting peace
The mode of full sensitizing range has following several:1) parameter configuration files of last time preservation are read;2) delimited by user by mouse;
3) set by user's input coordinate value.
A kind of 3. intrusion detection method that crosses the border in security sensitive region according to claim 1, it is characterised in that the back of the body
The one kind of scape relief method in mixed Gauss model method, VIBE algorithms, SACON algorithms, PBAS algorithms.
4. the intrusion detection method that crosses the border in a kind of security sensitive region according to claim 3, it is characterised in that use
PBAS algorithms extract moving target information from video flowing, and detailed process includes:
2.1) background modeling
Initial value of the pixel value of N two field pictures as background model before collection:
B(xi)={ B1(xi),…,Bk(xi),…,BN(xi)}
Wherein, B (xi) represent pixel xiIn the background model at current time, Bk(xi) represent pixel xiThe pixel value at history k moment;
2.2) foreground detection
For any one pixel x in current image framei, judge that formula is:
<mrow>
<mi>F</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mn>1</mn>
</mtd>
<mtd>
<mrow>
<mi>n</mi>
<mi>u</mi>
<mi>m</mi>
<mo>{</mo>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>&lsqb;</mo>
<mi>I</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>,</mo>
<msub>
<mi>B</mi>
<mi>k</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
<mo><</mo>
<mi>R</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo><</mo>
<mo>#</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mn>0</mn>
</mtd>
<mtd>
<mrow>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein, F (xi)=1 represents xiIt is judged as foreground point, F (xi)=0 represents xiIt is judged as background dot, num (.) represent full
The number of condition in sufficient bracket, dist (.) represent the Euclidean distance between 2 points in bracket;
2.3) context update
Background model is updated with the pixel for being identified as background, context update is divided into following three steps:
2.3.1) renewal background model:From its background model B (xi) one pixel value of middle random selection, with I (xi) replace, it is replaced
It is p (x to change probabilityi)=1/T (xi), T (xi) represent context update rate;
2.3.2 domain background model) is updated:From xiField in randomly choose pixel yi, with its current pixel value V (yi) replace
Change its background model B (yi) one pixel value of middle random selection, it is p (x that it, which replaces probability,i)=1/T (xi)。
2.3.3 judgment threshold R (x) are adaptively adjustedi) and turnover rate T (xi):Calculate pixel x in N two field pictures in the pastiIt is corresponding
The average value of background sample minimum range represents background complexity, and prospect judgment threshold is adjusted according to background complicacy self-adaptive
R(xi) and background model turnover rate T (xi)。
A kind of 5. intrusion detection method that crosses the border in security sensitive region according to claim 1, it is characterised in that the profit
The foreground image come out with morphologic filtering method processing detection, it is specific as follows:Reflation is first corroded by opening operation, removed
The profile of small noise and smooth motion object;First expanded by closed operation and corroded again, tiny sky inside filling foreground target
Hole.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710833558.1A CN107705326A (en) | 2017-09-15 | 2017-09-15 | A kind of intrusion detection method that crosses the border in security sensitive region |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710833558.1A CN107705326A (en) | 2017-09-15 | 2017-09-15 | A kind of intrusion detection method that crosses the border in security sensitive region |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107705326A true CN107705326A (en) | 2018-02-16 |
Family
ID=61172662
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710833558.1A Pending CN107705326A (en) | 2017-09-15 | 2017-09-15 | A kind of intrusion detection method that crosses the border in security sensitive region |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107705326A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110427812A (en) * | 2019-06-21 | 2019-11-08 | 武汉倍特威视***有限公司 | Colliery industry driving not pedestrian detection method based on video stream data |
CN111866589A (en) * | 2019-05-20 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Video data verification method and device, electronic equipment and storage medium |
CN111986378A (en) * | 2020-07-30 | 2020-11-24 | 湖南长城信息金融设备有限责任公司 | Bill color fiber yarn detection method and system |
CN112333417A (en) * | 2020-04-30 | 2021-02-05 | 深圳Tcl新技术有限公司 | Doorbell disturbance-free mode starting method and device and computer-readable storage medium |
CN113344874A (en) * | 2021-06-04 | 2021-09-03 | 温州大学 | Pedestrian boundary crossing detection method based on Gaussian mixture modeling |
CN113449675A (en) * | 2021-07-12 | 2021-09-28 | 西安科技大学 | Coal mine personnel border crossing detection method |
CN117079219A (en) * | 2023-10-08 | 2023-11-17 | 山东车拖车网络科技有限公司 | Vehicle running condition monitoring method and device applied to trailer service |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040004659A1 (en) * | 2002-07-02 | 2004-01-08 | Foote Jonathan T. | Intersection detection in panoramic video |
CN101236606A (en) * | 2008-03-07 | 2008-08-06 | 北京中星微电子有限公司 | Shadow cancelling method and system in vision frequency monitoring |
CN101339688A (en) * | 2008-08-27 | 2009-01-07 | 北京中星微电子有限公司 | Intrusion checking method and system |
CN104680555A (en) * | 2015-02-13 | 2015-06-03 | 电子科技大学 | Border-crossing detection method and border-crossing monitoring system based on video monitoring |
CN104700430A (en) * | 2014-10-05 | 2015-06-10 | 安徽工程大学 | Method for detecting movement of airborne displays |
CN103456028B (en) * | 2013-08-30 | 2016-08-31 | 浙江立元通信技术有限公司 | A kind of moving target detecting method |
-
2017
- 2017-09-15 CN CN201710833558.1A patent/CN107705326A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040004659A1 (en) * | 2002-07-02 | 2004-01-08 | Foote Jonathan T. | Intersection detection in panoramic video |
CN101236606A (en) * | 2008-03-07 | 2008-08-06 | 北京中星微电子有限公司 | Shadow cancelling method and system in vision frequency monitoring |
CN101339688A (en) * | 2008-08-27 | 2009-01-07 | 北京中星微电子有限公司 | Intrusion checking method and system |
CN103456028B (en) * | 2013-08-30 | 2016-08-31 | 浙江立元通信技术有限公司 | A kind of moving target detecting method |
CN104700430A (en) * | 2014-10-05 | 2015-06-10 | 安徽工程大学 | Method for detecting movement of airborne displays |
CN104680555A (en) * | 2015-02-13 | 2015-06-03 | 电子科技大学 | Border-crossing detection method and border-crossing monitoring system based on video monitoring |
Non-Patent Citations (2)
Title |
---|
于瑞云: "《计算几何及应用》", 30 April 2014, 哈尔滨工业大学出版社 * |
陈星明: "基于背景建模的运动目标监控视频检测算法", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111866589A (en) * | 2019-05-20 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Video data verification method and device, electronic equipment and storage medium |
CN110427812A (en) * | 2019-06-21 | 2019-11-08 | 武汉倍特威视***有限公司 | Colliery industry driving not pedestrian detection method based on video stream data |
CN112333417A (en) * | 2020-04-30 | 2021-02-05 | 深圳Tcl新技术有限公司 | Doorbell disturbance-free mode starting method and device and computer-readable storage medium |
CN111986378A (en) * | 2020-07-30 | 2020-11-24 | 湖南长城信息金融设备有限责任公司 | Bill color fiber yarn detection method and system |
CN113344874A (en) * | 2021-06-04 | 2021-09-03 | 温州大学 | Pedestrian boundary crossing detection method based on Gaussian mixture modeling |
CN113344874B (en) * | 2021-06-04 | 2024-02-09 | 温州大学 | Pedestrian boundary crossing detection method based on Gaussian mixture modeling |
CN113449675A (en) * | 2021-07-12 | 2021-09-28 | 西安科技大学 | Coal mine personnel border crossing detection method |
CN113449675B (en) * | 2021-07-12 | 2024-03-29 | 西安科技大学 | Method for detecting crossing of coal mine personnel |
CN117079219A (en) * | 2023-10-08 | 2023-11-17 | 山东车拖车网络科技有限公司 | Vehicle running condition monitoring method and device applied to trailer service |
CN117079219B (en) * | 2023-10-08 | 2024-01-09 | 山东车拖车网络科技有限公司 | Vehicle running condition monitoring method and device applied to trailer service |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107705326A (en) | A kind of intrusion detection method that crosses the border in security sensitive region | |
US20240062518A1 (en) | Method, apparatus, computer device and storage medium for detecting objects thrown from height | |
CN104680555B (en) | Cross the border detection method and out-of-range monitoring system based on video monitoring | |
CN108040221B (en) | Intelligent video analysis and monitoring system | |
CN104303193B (en) | Target classification based on cluster | |
CN105426820B (en) | More people's anomaly detection methods based on safety monitoring video data | |
US10366509B2 (en) | Setting different background model sensitivities by user defined regions and background filters | |
JP4668978B2 (en) | Flame detection method and apparatus | |
CN110428522A (en) | A kind of intelligent safety and defence system of wisdom new city | |
CN107657244B (en) | Human body falling behavior detection system based on multiple cameras and detection method thereof | |
CN111160125A (en) | Railway foreign matter intrusion detection method based on railway monitoring | |
CN103839085B (en) | A kind of detection method of compartment exception crowd density | |
CN108965826A (en) | Monitoring method, device, processing equipment and storage medium | |
CN107948465A (en) | A kind of method and apparatus for detecting camera and being disturbed | |
CN106228709B (en) | A kind of wisdom gold eyeball identifies that one adds paper money alarm method and device | |
WO2022078182A1 (en) | Throwing position acquisition method and apparatus, computer device and storage medium | |
CN101615295A (en) | Image processing system, image processing method and computer program | |
CN110636281B (en) | Real-time monitoring camera shielding detection method based on background model | |
CN111507235B (en) | Railway perimeter foreign matter intrusion detection method based on video | |
CN103929592A (en) | All-dimensional intelligent monitoring equipment and method | |
CN106682619A (en) | Object tracking method and device | |
CN111325051A (en) | Face recognition method and device based on face image ROI selection | |
CN106127814A (en) | A kind of wisdom gold eyeball identification gathering of people is fought alarm method and device | |
CN110255318B (en) | Method for detecting idle articles in elevator car based on image semantic segmentation | |
CN109377713A (en) | A kind of fire alarm method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180216 |