CN101303727B - Intelligent management method based on video human number Stat. and system thereof - Google Patents

Intelligent management method based on video human number Stat. and system thereof Download PDF

Info

Publication number
CN101303727B
CN101303727B CN2008101163123A CN200810116312A CN101303727B CN 101303727 B CN101303727 B CN 101303727B CN 2008101163123 A CN2008101163123 A CN 2008101163123A CN 200810116312 A CN200810116312 A CN 200810116312A CN 101303727 B CN101303727 B CN 101303727B
Authority
CN
China
Prior art keywords
video
intelligent management
module
demographics
people
Prior art date
Application number
CN2008101163123A
Other languages
Chinese (zh)
Other versions
CN101303727A (en
Inventor
卢晓鹏
王磊
Original Assignee
北京中星微电子有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 北京中星微电子有限公司 filed Critical 北京中星微电子有限公司
Priority to CN2008101163123A priority Critical patent/CN101303727B/en
Publication of CN101303727A publication Critical patent/CN101303727A/en
Application granted granted Critical
Publication of CN101303727B publication Critical patent/CN101303727B/en

Links

Abstract

The invention relates to the technical field of video image processing and mode identifying. The invention discloses an intelligent management method based on video people number statistics, which includes the following steps: S1: capturing a video flow image as an input image; S2: carrying out the treatment of people number statistics on the input image to obtain the people flow distribution data in the unit time of each area; S3: a control center carries out adjusting and control on the working personnel according to the people flow distribution data obtained in step S2. A corresponding system includes a video collection module and a people number detecting statistics module; wherein, the people number detecting statistics module include a background construction module, a motion detecting module, an area analyzing module and a data statistics module. The method can be used for reasonably arranging the working number of the personnel of different posts according to the people flow statistic in a scene in a unit time, can better and more reasonably save the manpower cost and effectively improve the management level of public areas.

Description

Intelligent management and system thereof based on the video demographics

Technical field

The present invention relates to video image and handle and mode identification technology, relate in particular to a kind of intelligent management and system thereof based on the video demographics.

Background technology

In today that the information system management level improves day by day, carry out volume of the flow of passengers statistics such as the volume of the flow of passengers is estimated in real time, passenger flow distributional analysis, degree of crowding estimation become provides first-hand background information for the public domain management effective way for the huge place of flows of the people such as supermarket, market, station, bank.Address this problem, rely on surveillance equipment and artificial judgment processing merely and be the needs that can not satisfy high-efficiency management far away, video image processing and data statistical approach and technology become the key that improves the public domain management level effectively.

Undertaken by the dynamic video data aspect the target statistics, the patent of invention that application number is 03109626.3, name is called " small insect automatic counter system " discloses the automatic counter system of a kind of small insect, this invention belongs to automatic measurement, counting field, automatic counter system is wherein operated according to following steps, 1. image acquisition, the unit equipment that utilizes digital camera or CCD camera and image pick-up card with the image acquisition of insect to and send into the computing machine input end; 2. utilize the computer programs process view data, cromogram is converted into gray-scale map; 3. gray-scale map is carried out Threshold Segmentation, become binary map; 4. binary map is carried out connected component labeling; 5. carrying out objective body identification and non-objective body filters; 6. count the number of the contained insect of connected region; 7. count results is shown to the user.

Yet, foregoing invention only is the automatic counting of realizing in specific background environment and specific region the specific objective body, and this counting technology at the specific objective body can not satisfy under the various different background environment in public domain the statistical demand to the flow of the people of continuous variation.

Summary of the invention

The present invention proposes in order to address the above problem just.

The purpose of this invention is to provide a kind of intelligent management and system based on the video demographics.

Described intelligent management based on the video demographics comprises the steps:

The video streaming image that at first obtains guarded region is as input picture.

Then this input picture is carried out motion detection, according to this motion detection result each moving region is analyzed again and number is estimated, thereby obtain stream of people's distributed data of each area unit time; What wherein, this regional analysis method adopted is Overlapping Calculation and similarity calculating method.And this similarity calculating method is, at the color characteristic coefficient of similarity During greater than second threshold value, assert that these detected any two moving regions are same target.Wherein, this second threshold value greater than 0 less than 1.And described color characteristic coefficient of similarity satisfies Wherein, Be respectively the color probability distribution of these any two moving regions, and this color probability distribution is satisfied,

q ^ u = C Σ i = 1 n N ( x i * ; μ 0 ; σ 0 ) δ [ b ( x i * ) - u ]

Wherein, I=1 ..., n is the set of human body target elliptic region picture point, μ 0Be human body target elliptical center, σ 0Be ellipse long and short shaft parameter matrix, function R 2→ 1 ..., m} is a pixel Color value, u is a color component discretize grade, C for the standardization constant, δ is a kronecker delta function, N () is a gaussian kernel function, m is a gray shade scale.

Last control center carries out staff's regulation and control according to the stream of people's distributed data that obtains among the step S2.

Described intelligent management system based on the video demographics comprises video acquisition module, number detection statistics module and control center.

This video acquisition module is used to obtain the video streaming image of guarded region, and with it as input picture.

This number detection statistics module is used for this input picture is carried out motion detection, and according to this motion detection result each moving region is analyzed and the number estimation, to obtain stream of people's distributed data of each area unit time.What wherein, this regional analysis adopted is Overlapping Calculation and similarity calculating method; And this similarity calculating method is, at the color characteristic coefficient of similarity During greater than second threshold value, assert that these detected any two moving regions are same target.Wherein, this second threshold value greater than 0 less than 1.And described color characteristic coefficient of similarity satisfies Wherein, Be respectively the color probability distribution of these any two moving regions, and this color probability distribution is satisfied,

q ^ u = C Σ i = 1 n N ( x i * ; μ 0 ; σ 0 ) δ [ b ( x i * ) - u ]

Wherein, I=1 ..., n is the set of human body target elliptic region picture point, μ 0Be human body target elliptical center, σ 0Be ellipse long and short shaft parameter matrix, function R 2→ 1 ..., m} is a pixel Color value, u is a color component discretize grade, C for the standardization constant, δ is a kronecker delta function, N () is a gaussian kernel function, m is a gray shade scale.

This control center carries out staff's regulation and control according to described stream of people's distributed data.

The present invention propose based on the intelligent management of video demographics and system can be applicable to that customer quantity under the multiple occasion such as supermarket, market, station, bank is estimated in real time, passenger flow distributional analysis, degree of crowding estimation etc., the important application of this type of intelligent video monitoring can improve the management level such as the public domain greatly.

Description of drawings

Fig. 1 represents system chart of the present invention;

Fig. 2 represents the structured flowchart of number detection statistics module of the present invention.

Embodiment

Following examples are used to illustrate the present invention, but are not used for limiting the scope of the invention.

The present invention is directed to general video monitoring scene, obtain moving region in the video flowing by the method for motion detection, on the basis of regional analysis, the human body number of each moving region and many people zone cut apart according to a preliminary estimate, adopt the method for target following that all moving regions in the visual field are followed the tracks of then, number in the accurate statistical picture, according to stream of people's statistic in the scene in the unit interval section, reasonably arrange employee's number on duty in different posies, better rationally save human cost, reduce client and wait in line the time.

Fig. 1 is a system chart of the present invention.As shown in the figure, system mainly comprises video acquisition module, number detection statistics module and control center.In addition, system also comprises employee's calling device, and cashier, warehouse and sales promotion etc.

The major function of video acquisition module is that video streaming image is taken and obtained to monitoring scene, can be by such as the monitoring camera of special use or traditional camera is taken and the capturing video stream picture realizes the function of this module.

The major function of number detection module is the number that detects the personnel in the input picture in the unit interval section, can adopt multiple existing motion detection technique, grade as background subtraction point-score, frame-to-frame differences point-score, mixed Gaussian background subtraction, video flowing according to input is exported the personnel's number statistical in the input picture by a series of data processing, and obtains stream of people's distributed data of each area unit time.

The video streaming image that obtains by IP Camera or video camera by number detection statistics module, obtains people's distributions of the unit interval in each location, supermarket respectively as input picture, distributes according to this, and control center carries out staff's regulation and control.Such as, control center sends information to the calling device that the employee carries, arrange it to arrive the nervous relatively department of work staff, as cashier of peak period etc., perhaps being arranged into the warehouse gets in stocks to shelf, guarantee goods abundance for sale on the shelf, perhaps the whereabouts client propagates the characteristics of commodity, carries out commercial promotions etc.

Accordingly, the intelligent management based on the video demographics of the present invention's proposition is realized as follows:

S1: obtain video streaming image as input picture;

S2: input picture is carried out demographics handle, obtain stream of people's distributed data of each area unit time;

S3: control center carries out staff's regulation and control according to the stream of people's distributed data that obtains among the step S2.

Fig. 2 represents the structured flowchart of number detection statistics module of the present invention.As shown in Figure 2, in the present invention, number detection statistics module comprises background constructing module, motion detection block, regional analysis module and data statistics module.

Wherein the background constructing module is used to construct the background model in zone to be monitored, and the simplest background model is the time average image, promptly utilizes Same Scene in the average image of the period background model as this scene.

In a preferred embodiment of the invention, the background constructing module is made up of background modeling module and context update module.Wherein, the background modeling module is used for drawing each pixel gray-scale value of initial background image according to the some frames from the collection of image acquisition equipment; Whether the absolute difference of the pixel gray-scale value that the context update module is used for judging that described present frame is identical with its former frame coordinate greater than preset threshold, if, then make B (x, y)=α B 1(x, y), if not, then make B (x, y)=α B 1(x, y)+(1-α) f (x, y).Wherein (x is that coordinate is (x, pixel gray-scale value y) in the background images of described present frame y) to B; If present frame is the 1st frame after some frames that image acquisition equipment is gathered in the background modeling module, then B 1(x is that coordinate is (x, pixel gray-scale value y), otherwise be that coordinate is (x, pixel gray-scale value y) in the background images of preceding 1 frame of described present frame in the described initial background image y); (x is that coordinate is that (x, pixel gray-scale value y), α are the parameters of setting, and value satisfies 0≤α≤1 in the described present frame y) to f.

In motion detection block, for current input image, at first it is obtained difference image with background images and former frame image subtraction respectively, use thresholding method that two width of cloth difference images are carried out binary conversion treatment respectively.Then, use Mathematical Morphology Method (such as dilation operation, erosion operation, opening operation, pass computing etc.) that two width of cloth binary images are carried out Filtering Processing, fill the cavity in the foreground area, remove the less isolated area of area, non-connected region simultaneously, only keep the connected component of the area of connected region greater than given threshold value.At last, above-mentioned two filtered binary images are carried out the logical and operation, and the image after the computing is carried out mathematical morphology filter handle, obtain final motion detection result.After obtaining motion detection result, non-moving region is upgraded according to the more new model of background subtraction point-score.

The regional analysis module functions is that to analyze each moving region be single zone or many people zone, if many people zone number in the estimation region and be a plurality of single zones with Region Segmentation then.

The present invention comes the realization number to estimate by identification people's position, the crown.Generally, people's crown point generally is visible in the video image, and we are in conjunction with the position that the geometric configuration and the vertical projection of moving region are determined the people crown in the scene, and the crowd is cut apart in number and then realization in the estimated image.At first, calculate the vertical projection in each two-value zone, find all Local Extremum of vertical projection.Then, investigate all Local Extremum,, then think possible crown point if the value of the projection of certain extreme point is bigger than a given threshold value.At last, all possible crown points are merged, the possible crown point that is about to close proximity on the horizontal direction is merged into a stature summit, obtains final crown point.

Realize also needing many people zone is cut apart after the number estimation in the zone.In a preferred embodiment of the present invention, adopt the oval method that fits to cut apart many people zone.If the ratio of the width of human body and height is a fixed value l, be end points with the crown point of human body, use the height of a major axis and human body identical, breadth length ratio is that the ellipse encirclement human region of l can obtain a single zone.

In the video monitoring process, human body as monitoring objective generally is kept in motion, therefore same target body may be in different positions in different video picture frame especially adjacent image frame, thereby solves the object matching problem of adjacent two interframe of video image with regard to needs.Adopt single regional matching module to solve the object matching problem of adjacent two interframe in the present invention.Wherein mainly adopt Overlapping Calculation and similarity Calculation Method to realize the coupling of target.

Overlapping Calculation

The present invention utilizes the overlap coefficient of adjacent two frames two target regions to judge whether two targets mate, and in general video monitoring system, the movement velocity of human body is not too large, and same human region overlapping degree in adjacent two frames is still very high.

If any two shared rectangles in zone are respectively R 1And R 2, calculate R 1And R 2Coincidence district R 1∩ R 2, then the overlap coefficient in these two zones calculates according to following formula:

r = s ( R 1 ∩ R 2 ) min ( s ( R 1 ) , s ( R 2 ) ) - - - ( 1 )

Wherein, s () is the operator of expression reference area.During actual computation, can set one as required greater than 0 less than 1 threshold value Th1, that can think when r>Yh1 that two zones have a coupling may.

The overlapping target association that carries out of simple dependence area, often not accurate enough, thus can be by more target signature, select for use color characteristic similarity coupling to come further target association in preferred an enforcement of the present invention.

Similarity is calculated

If the set of human body target elliptic region picture point is I=1 ..., n, the center is μ 0, ellipse long and short shaft parameter matrix σ 0, wherein major axis parameter h and minor axis parameter w, defined function R 2→ 1 ..., the m} remarked pixel Color value.Then in the ellipse target zone, picture point The color probability distribution U=1 ..., m. can be expressed as:

q ^ u = C Σ i = 1 n N ( x i * ; μ 0 ; σ 0 ) δ [ b ( x i * ) - u ] - - - ( 11 )

Wherein, δ is a kronecker delta function, and C is the standardization constant, and has N () is a gaussian kernel function, and u is a color component discretize grade.

Suppose that two degrees of overlapping are respectively Q greater than the target body zone of thresholding 1And Q 2, both color probability distribution are respectively With Their similarity can be measured with the Bhattacharrya coefficient, that is:

Set one greater than 0 less than 1 threshold value Th2, when ρ>Th2, can think two target area associations to be same target.

Data statistics module is set up a human region cdr database, mainly be responsible for the size in certain scene in the record unit time section, position, characteristics of image and should the zone in number, use for the later stage and to handle use.

Because because the gateway in supermarket has unidirectional characteristics, therefore in a specific embodiment of the present invention, the regional analysis module is only carried out demographics in the starting stage to entire image, the statistics in later stage is only carried out the borderline region of image, auxiliary with the strong method of engineerings such as Kalman filtering and template matches, according to the number that obtains passing in and out scene, derive personnel's number of whole scene.

N all=N t-1+N in-N out

Wherein, N In, N OutBe respectively personnel's number of unit interval turnover scene.

Intelligent management system based on the video demographics according to the present invention is analyzed the video image in the monitoring scene based on computer vision technique, statistics is extracted the number information in the scene, personnel's the work of reasonably arranging work distributes, for saving human cost, making things convenient for customers, reduce the stream of people's irrational situations such as queuing time is permanent of crowding, avoid paying the bill active and effective effect is arranged.

Compared with prior art, remarkable advantage of the present invention comprises:

1. the intelligent management system integrated level height based on the video demographics of the present invention's proposition is easy to use, is easy to installation and maintenance.

2. the intelligent management system based on the video demographics of the present invention's proposition is embedded in front end with context update, number detection and analysis, only sends the number statistical value to control center, occupied bandwidth is little, required storage space is little, and real-time is good, system effectiveness height, stability.

3. the demographics detection module clear in structure in the system, the each several part division of labor is clear and definite, independence is strong, calculates easy, speed fast, be easy to the hardware realization.

4. the control center's module in the system of the present invention is carefully thought out, and better rationally saves human cost, and shopping makes things convenient for customers.

5. in addition, the method that combining target of the present invention is followed the tracks of is optimized the method for single frames demographics, has further improved the accuracy of demographics.

Though the present invention explains in conjunction with an embodiment; but those skilled in the art can be to wherein some feature appropriate change or apply it to other field addressing the above problem in addition, so all relevant expansions of carrying out on the basis of present embodiment of those skilled in the art and use the protection domain that all should fall into the application.

Claims (14)

1. the intelligent management based on the video demographics is characterized in that this method comprises the steps:
S1: the video streaming image that obtains guarded region is as input picture;
S2: this input picture is carried out motion detection, according to this motion detection result each moving region is analyzed again and number is estimated, thereby obtain stream of people's distributed data of each area unit time; What wherein, this regional analysis method adopted is Overlapping Calculation and similarity calculating method; And this similarity calculating method is, at the color characteristic coefficient of similarity During greater than second threshold value, assert that these detected any two moving regions are same target; Wherein, this second threshold value greater than 0 less than 1;
And described color characteristic coefficient of similarity satisfies Wherein, Be respectively the color probability distribution of these any two moving regions, and this color probability distribution is satisfied,
q ^ u = C Σ i = 1 n N ( x i * ; μ 0 ; σ 0 ) δ [ b ( x i * ) - u ]
Wherein, I=1 ..., n is the set of human body target elliptic region picture point, μ 0Be human body target elliptical center, σ 0Be ellipse long and short shaft parameter matrix, function R 2→ 1 ..., m} is a pixel Color value, u is a color component discretize grade, C for the standardization constant, δ is a kronecker delta function, N () is a gaussian kernel function, m is a gray shade scale;
S3: control center carries out staff's regulation and control according to the stream of people's distributed data that obtains among the step S2.
2. the intelligent management based on the video demographics as claimed in claim 1 is characterized in that in step S2,
At first construct the background model in zone to be monitored, thereby according to this background model that obtains described input picture is carried out motion detection then.
3. the intelligent management based on the video demographics as claimed in claim 2 is characterized in that in the process of the background model that makes up zone to be monitored, background is carried out timing upgrade.
4. the intelligent management based on the video demographics as claimed in claim 1 is characterized in that in motion detection process, adopts the method for difference image, binary conversion treatment and Filtering Processing to obtain final motion detection result.
5. the intelligent management based on the video demographics as claimed in claim 1, it is characterized in that in the process that each moving region is analyzed, at first analyzing each moving region is single zone or many people zone, estimates the number in many people zone then and it is divided into a plurality of single zones.
6. the intelligent management based on the video demographics as claimed in claim 5 is characterized in that when many people zone is cut apart, and adopts the oval method that fits to cut apart many people zone.
7. the intelligent management based on the video demographics as claimed in claim 1, it is characterized in that described Overlapping Calculation method is, during greater than first threshold, assert that these detected any two moving regions are complementary at degree of overlapping coefficient r, and this first threshold greater than 0 less than 1;
Described degree of overlapping coefficient r satisfies Wherein, R 1And R 2Be respectively shared rectangular area, described two moving regions, R 1∩ R 2Be R 1And R 2The coincidence district, s () is the operator of reference area.
8. the intelligent management system based on the video demographics is characterized in that this system comprises video acquisition module, number detection statistics module and control center;
Described video acquisition module is used to obtain the video streaming image of guarded region, and with it as input picture;
Described number detection statistics module is used for this input picture is carried out motion detection, and according to this motion detection result each moving region is analyzed and the number estimation, to obtain stream of people's distributed data of each area unit time; What wherein, this regional analysis adopted is Overlapping Calculation and similarity calculating method; And this similarity calculating method is, at the color characteristic coefficient of similarity During greater than second threshold value, assert that these detected any two moving regions are same target; Wherein, this second threshold value greater than 0 less than 1;
And described color characteristic coefficient of similarity satisfies Wherein, Be respectively the color probability distribution of these any two moving regions, and this color probability distribution is satisfied,
q ^ u = C Σ i = 1 n N ( x i * ; μ 0 ; σ 0 ) δ [ b ( x i * ) - u ]
Wherein, I=1 ..., n is the set of human body target elliptic region picture point, μ 0Be human body target elliptical center, σ 0Be ellipse long and short shaft parameter matrix, function R 2→ 1 ..., m} is a pixel Color value, u is a color component discretize grade, C for the standardization constant, δ is a kronecker delta function, N () is a gaussian kernel function, m is a gray shade scale;
Described control center carries out staff's regulation and control according to described stream of people's distributed data.
9. the intelligent management system based on the video demographics as claimed in claim 8 is characterized in that described number detection statistics module comprises:
The background constructing module is used to construct the background model in zone to be monitored;
Motion detection block, thus according to this background model described input picture is carried out motion detection.
10. the intelligent management system based on the video demographics as claimed in claim 9, it is characterized in that described background constructing module comprises that being used for basis draws the background modeling module of each pixel gray-scale value of initial background image from some frames of video acquisition module collection, and be used for background is carried out the regularly context update module of renewal.
11. the intelligent management system based on the video demographics as claimed in claim 9 is characterized in that described motion detection block adopts the method for difference image, binary conversion treatment and Filtering Processing to obtain final motion detection result.
12. the intelligent management system based on the video demographics as claimed in claim 8, it is characterized in that, in the described number detection statistics module each moving region being analyzed is that to analyze each moving region be single zone or many people zone, and estimates the number in many people zone and it is divided into a plurality of single zones.
13. the intelligent management system based on the video demographics as claimed in claim 12, it is characterized in that described Overlapping Calculation method is at degree of overlapping coefficient r during greater than first threshold, assert that these detected any two moving regions are complementary, and this first threshold greater than 0 less than 1;
Described degree of overlapping coefficient r satisfies Wherein, R 1And R 2Be respectively shared rectangular area, described two moving regions, R 1∩ R 2Be R 1And R 2The coincidence district, s () is the operator of reference area.
14. the intelligent management system based on the video demographics as claimed in claim 8, it is characterized in that, it is in the starting stage entire image to be carried out demographics that described number detection statistics module is analyzed each moving region, statistics in the later stage is only carried out the borderline region of image, auxiliary with Kalman filtering and the strong method of template matches engineering, according to the number that obtains passing in and out scene, derive personnel's number of whole scene.
CN2008101163123A 2008-07-08 2008-07-08 Intelligent management method based on video human number Stat. and system thereof CN101303727B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2008101163123A CN101303727B (en) 2008-07-08 2008-07-08 Intelligent management method based on video human number Stat. and system thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2008101163123A CN101303727B (en) 2008-07-08 2008-07-08 Intelligent management method based on video human number Stat. and system thereof

Publications (2)

Publication Number Publication Date
CN101303727A CN101303727A (en) 2008-11-12
CN101303727B true CN101303727B (en) 2011-11-23

Family

ID=40113626

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2008101163123A CN101303727B (en) 2008-07-08 2008-07-08 Intelligent management method based on video human number Stat. and system thereof

Country Status (1)

Country Link
CN (1) CN101303727B (en)

Families Citing this family (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101420536B (en) * 2008-11-28 2011-03-02 江苏科海智能系统有限公司 Background image modeling method for video stream
CN101577812B (en) * 2009-03-06 2014-07-30 北京中星微电子有限公司 Method and system for post monitoring
CN102098491B (en) * 2009-12-14 2014-05-21 北京中星微电子有限公司 Monitoring method and device of chaotic scene
GB2476500B (en) * 2009-12-24 2012-06-20 Infrared Integrated Syst Ltd Activity mapping system
CN102063753A (en) * 2011-01-18 2011-05-18 王芳 Method for identification and statistic of visitors based on Internet of things
CN102496164B (en) * 2011-11-14 2013-12-11 通号通信信息集团有限公司 Event detection method and event detection system
CN102542289B (en) * 2011-12-16 2014-06-04 重庆邮电大学 Pedestrian volume statistical method based on plurality of Gaussian counting models
CN102609229B (en) * 2012-02-16 2014-12-31 中国科学院声学研究所 Graph parallel-updating method for embedded multi-window application
CN104376575B (en) * 2013-08-15 2018-02-16 汉王科技股份有限公司 A kind of pedestrian counting method and device based on multi-cam monitoring
CN104392522B (en) * 2014-11-21 2016-08-24 安徽清新互联信息科技有限公司 A kind of passenger flow statistical system based on pseudo-depth information
CN104408746B (en) * 2014-11-21 2017-07-14 安徽清新互联信息科技有限公司 A kind of passenger flow statistical system based on depth information
CN106295788B (en) 2015-05-12 2019-01-08 杭州海康威视数字技术股份有限公司 The statistical method and device of the volume of the flow of passengers
CN105466414B (en) * 2015-11-17 2018-05-29 广东欧珀移动通信有限公司 The recommendation method and server of a kind of navigation routine
CN105574499B (en) * 2015-12-15 2019-08-20 东华大学 A kind of number detection statistics method and system based on SOC
CN107025418A (en) * 2016-01-29 2017-08-08 广州地铁集团有限公司 Anti- detection method of stealing a ride, detecting system and its system based on image procossing
CN105722271B (en) * 2016-02-04 2017-06-06 广东金南方新科技股份有限公司 Intelligent automatic radiating electric control box
CN105489017B (en) * 2016-02-04 2016-12-07 深圳亿维锐创科技股份有限公司 Multifunctional traffic crossing electrical control cabinet
CN106304564A (en) * 2016-09-28 2017-01-04 广东雅达电子股份有限公司 A kind of wisdom Lighting Control Assembly based on image recognition and method
CN107481255B (en) * 2017-08-08 2019-12-20 浙江大华技术股份有限公司 Method and device for determining number of people
CN107292297A (en) * 2017-08-09 2017-10-24 电子科技大学 A kind of video car flow quantity measuring method tracked based on deep learning and Duplication
CN107368823A (en) * 2017-08-23 2017-11-21 广州市九安光电技术股份有限公司 A kind of stream of people's focus monitoring method and system based on panoramic picture
CN107680118B (en) * 2017-08-29 2020-10-20 苏州佳世达光电有限公司 Image identification tracking method
CN108345841B (en) * 2018-01-23 2019-02-12 杭州视在科技有限公司 A kind of intelligent filtering method of video image processing
CN108921983B (en) * 2018-05-22 2020-11-03 浙江大学山东工业技术研究院 People flow statistical method for waiting room of complex equipment coal mine system
CN111290001A (en) * 2018-12-06 2020-06-16 杭州海康威视数字技术股份有限公司 Target overall planning method, device and equipment based on GPS coordinates
CN109920104A (en) * 2019-01-18 2019-06-21 浙江万邦智能科技股份有限公司 A kind of access control system based on recognition of face for wisdom hospital
CN110488695A (en) * 2019-08-09 2019-11-22 广州启盟信息科技有限公司 A kind of dispatching method and device based on stream of people's data

Also Published As

Publication number Publication date
CN101303727A (en) 2008-11-12

Similar Documents

Publication Publication Date Title
CN105447458B (en) A kind of large-scale crowd video analytic system and method
CN105069472B (en) A kind of vehicle checking method adaptive based on convolutional neural networks
Hadi et al. Vehicle detection and tracking techniques: a concise review
Hsieh et al. Shadow elimination for effective moving object detection by Gaussian shadow modeling
Pai et al. Pedestrian detection and tracking at crossroads
Ozkurt et al. Automatic traffic density estimation and vehicle classification for traffic surveillance systems using neural networks
KR101375583B1 (en) Object Density Estimation in Video
CN101799876B (en) Video/audio intelligent analysis management control system
CN101120382B (en) Method for tracking moving object in video acquired of scene with camera
CN100593335C (en) Motion detection method and device
CN102081918B (en) Video image display control method and video image display device
CN102214291B (en) Method for quickly and accurately detecting and tracking human face based on video sequence
Bas et al. Automatic vehicle counting from video for traffic flow analysis
CN104933710B (en) Based on the shop stream of people track intelligent analysis method under monitor video
CN102165493B (en) Detection of vehicles in an image
CN101739686B (en) Moving object tracking method and system thereof
US9159228B2 (en) System and method for estimation of available parking space through intersection traffic counting
Buch et al. Detection and classification of vehicles for urban traffic scenes
EP1697901B1 (en) Method for modeling background and foreground regions
Buch et al. Urban road user detection and classification using 3D wire frame models
CN101739551B (en) Method and system for identifying moving objects
CN101777114B (en) Intelligent analysis system and intelligent analysis method for video monitoring, and system and method for detecting and tracking head and shoulder
CN102163290B (en) Method for modeling abnormal events in multi-visual angle video monitoring based on temporal-spatial correlation information
Chiu et al. A robust object segmentation system using a probability-based background extraction algorithm
JP2008544705A (en) Detect and track surveillance objects from overhead video streams

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model
TR01 Transfer of patent right

Effective date of registration: 20160516

Address after: 519031 Guangdong city of Zhuhai province Hengqin Baohua Road No. 6, room 105 -478

Patentee after: GUANGDONG ZHONGXING ELECTRONICS CO., LTD.

Address before: 100083 Haidian District, Xueyuan Road, No. 35, the world building, the second floor of the building on the ground floor, No. 16

Patentee before: Beijing Vimicro Corporation

C41 Transfer of patent application or patent right or utility model