CN102184409A - Machine-vision-based passenger flow statistics method and system - Google Patents
Machine-vision-based passenger flow statistics method and system Download PDFInfo
- Publication number
- CN102184409A CN102184409A CN2011101016376A CN201110101637A CN102184409A CN 102184409 A CN102184409 A CN 102184409A CN 2011101016376 A CN2011101016376 A CN 2011101016376A CN 201110101637 A CN201110101637 A CN 201110101637A CN 102184409 A CN102184409 A CN 102184409A
- Authority
- CN
- China
- Prior art keywords
- data
- passenger flow
- people
- client
- model
- 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
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a machine-vision-based passenger flow statistics method and system. Based on the machine vision theory, the method comprises the following steps in sequence: building a model; extracting features of a target image for detection according to the model; and finally comparing all images, serially connecting the tracking points of all frames to completely confirm the track of one person at a fixed region, and obtaining the number of persons passing by the region through calculation of the number of the tracks. The statistics passenger flow is recognized from the acquired video images, the tracking objects are individual people, thus the possibility of interference by the crowded or other objects is small, thereby achieving the counting of persons in a monitored region under complicated conditions such as bidirectional movement and the like, and achieving the function of accurately counting the number of passengers.
Description
Technical field
The present invention relates to computer vision and image processing techniques neck, particularly a kind of method and system of carrying out passenger flow statistics based on machine vision.
Background technology
In the society of present information high speed development, different industries all has the demand of statistical number of person, and for example in the Operation Decision and integrated management of large scale business system, customer data accurately and timely plays important effect to steady field, Wang Chang; In the garden management system, visitor's data can provide important reference for Destination Management; In public security system, in time notifier's flow is unusual, thereby reduces potential safety hazard or the like.
Currently used demographics mainly contains following several mode: complicate statistics, infrared correlation statistics, video background modeling statistics.These statisticals have following problem:
The complicate statistics accuracy rate is the highest, but because excessive cost of labor makes the sampling that can only do the short time under most of situation, its shortcoming is apparent.
Infrared correlation statistics is equivalent to infrared injection device be made an invisible line, the people rushes across then count off, this method can only be used for the very little stream of people, the behavior that the very important person has and drains, then system can't calculate, be difficult to that competent many people enter side by side, personnel's two-way flow, to the statistics under the situations such as row enter in single file.
The video background modeling pattern uses video analysis to do the number statistics, and to recently detecting the object that moves, it utilizes the difference between image frame to set up background model by background, filters out the moving line that foreground object is also determined object.But use this technology an important prerequisite is arranged: require scene spaciousness, open, as frontier defense line etc.This is because object is crowded or when blocking, the prospect of different objects will can't be distinguished owing to inter-adhesive, and based on the also inefficacy simultaneously of estimation of prospect, crowded scene also can influence the correct foundation of background model, and under spacious scene, these rough sledding are all rare.But, the actual scene of most passenger flow statisticses does not satisfy " the spacious scene " of this technology and supposes, therefore use the method that two characteristics are arranged: the first, require video camera to lay vertically downward, this is because this class algorithm blocks highstrung cause to object.The second, along with the degree of crowding raising of passenger flow, the people that very difficult explanation is close together, the accuracy of passenger flow statistics will be subjected to obvious influence.
Summary of the invention
The purpose of this invention is to provide a kind of passenger flow statistical method and system, can realize accurate, adaptable passenger flow demographics function based on machine vision.
The objective of the invention is to be achieved through the following technical solutions:
A kind of passenger flow statistical method based on machine vision comprises that the following step poly-:
(1) sets up model,, obtain abundant people's picture sample, go out the model that feature is set up the people based on shape and color extracting by these samples at the scene of a fixed angle;
(2) extract feature according to model at target image and detect, in every frame video pictures, orient everyone position, people and other object areas are separated according to the single frames picture;
(3) the proprietary feature that obtains in everyone and the next frame image that relatively obtains in the previous frame image, two people that feature is the most close confirm as same individual, relatively more all images, the trace point of all frames is together in series people of complete affirmation at the track of this fixed area, can draws by this regional number by the bar number that calculates track.
The data of described model 90% are extracted based on shape, and 10% data are based on color extracting.
When calculating by this regional number, everyone enters into or walks out this zone, realizes by the direction of calculating track.
Based on above-mentioned statistical method, a kind of realization system is provided, forms by video camera, statistics service end, data server, client, data server and client composition LAN (Local Area Network),
Described video camera carries out the data acquisition in field monitor zone;
Described statistics service end: the video processing part of system, be responsible for calculating the real-time discrepancy data of correspondence image;
Described data server: Web server, the controller of system is used for receiving and dispatching order, coordinates statistics service end and client, uses the PostGreSql database, and the data of reception client are also stored;
Described client: be used for carrying out data query by web.
Passenger flow statistical method and system based on machine vision of the present invention, based on the machine vision theory, the identification statistics volume of the flow of passengers from the video image of gathering, follow the tracks of to as if independent one by one people, and be not vulnerable to the interference of crowded or other objects, the enumeration problem under the complex situations such as personnel's bidirectional-movement in the guarded region can be overcome, accurate passenger flow demographics function can be realized.
Embodiment
As shown in Figure 1, provide a kind of system of passenger flow statistical method of the present invention that realizes and formed structural drawing, it is made up of video camera, statistics service end, data server, client, and data server and client are formed LAN (Local Area Network), and described video camera carries out the data acquisition in field monitor zone; Described statistics service end: the video processing part of system, be responsible for calculating the real-time discrepancy data of correspondence image; Described data server: Web server, the controller of system is used for receiving and dispatching order, coordinates statistics service end and client, uses the PostGreSq1 database, and the data of reception client are also stored; Described client: be used for carrying out data query by web.
With a business consortium is example, and this system has each independent market to gather people's logarithmic data, uploads to general headquarters of group, finishes data and gathers, and can carry out the part inquiry and also build server in each market.
The video camera that front end is installed passes to the statistics service end to signal by concentric cable, the polylith image pick-up card is housed on each service end, image pick-up card is responsible for finishing the analog signal conversion of image, each number statistical module all is connected with the output of a certain road capture card, is responsible for calculating the real-time discrepancy data of correspondence image.
The demographics module adopts the passenger flow statistical method based on machine vision, comprises that the following step poly-:
(1) sets up model,, obtain abundant people's picture sample, go out the model that feature is set up the people based on shape and color extracting by these samples at the scene of a fixed angle;
(2) extract feature according to model at target image and detect, in every frame video pictures, orient everyone position, people and other object areas are separated according to the single frames picture;
(3) the proprietary feature that obtains in everyone and the next frame image that relatively obtains in the previous frame image, two people that feature is the most close confirm as same individual, relatively more all images, the trace point of all frames is together in series people of complete affirmation at the track of this fixed area, can draws by this regional number by the bar number that calculates track.
The data of described model 90% are extracted based on shape, and 10% data are based on color extracting.
When calculating by this regional number, everyone enters into or walks out this zone, realizes by the direction of calculating track.
The every some minutes of each number statistical module produces the record of the number of coming in and going out, and represents the personnel amount of turnover module region within the jurisdiction during this, and these records are called as raw readings, and system guarantees that they can be distorted in no instance.Each bar raw readings is duplicated into two parts, and a copy of it sends to data server, and another part then directly deposits the service end local backup in.Make when data server goes wrong that the demographics module of statistics on the service end still can operate as normal, server recover normal after, the data of delay will be extracted from the backup of statistics service end and resend to data server.
Data server receives and respectively adds up the data recording that service end is sent, and deposits it in its central database.It is regularly safeguarded automatically to central database, checks and the revision abnormal data, generates conventional form.
Client is the main mode of user's manipulation data server, by the ruuning situation that its user can understand service end at any time, monitors each regional real time data, the historical data of inquiry or revision central database.Especially, client can also provide the real-time video picture of appointed area to the user.In addition, the configuration to system's types of functionality also can realize by monitor client.
Claims (4)
1. passenger flow statistical method based on machine vision is characterized in that comprising that the following step poly-:
(1) sets up model,, obtain abundant people's picture sample, go out the model that feature is set up the people based on shape and color extracting by these samples at the scene of a fixed angle;
(2) extract feature according to model at target image and detect, in every frame video pictures, orient everyone position, people and other object areas are separated according to the single frames picture;
(3) the proprietary feature that obtains in everyone and the next frame image that relatively obtains in the previous frame image, two people that feature is the most close confirm as same individual, relatively more all images, the trace point of all frames is together in series people of complete affirmation at the track of this fixed area, can draws by this regional number by the bar number that calculates track.
2. the passenger flow statistical method based on machine vision according to claim 1 is characterized in that, the data of described model 90% are extracted based on shape, and 10% data are based on color extracting.
3. the passenger flow statistical method based on machine vision according to claim 1 is characterized in that, when calculating by this regional number, everyone enters into or walk out this zone, realizes by the direction of calculating track.
4. a system that adopts the described method of claim 1 to realize passenger flow statistics is characterized in that, is made up of video camera, statistics service end, data server, client, and data server and client are formed LAN (Local Area Network),
Described video camera carries out the data acquisition in field monitor zone;
Described statistics service end: the video processing part of system, be responsible for calculating the real-time discrepancy data of correspondence image;
Described data server: Web server, the controller of system is used for receiving and dispatching order, coordinates statistics service end and client, uses the PostGreSq1 database, and the data of reception client are also stored; Described client: be used for carrying out data query by web.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011101016376A CN102184409A (en) | 2011-04-22 | 2011-04-22 | Machine-vision-based passenger flow statistics method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011101016376A CN102184409A (en) | 2011-04-22 | 2011-04-22 | Machine-vision-based passenger flow statistics method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102184409A true CN102184409A (en) | 2011-09-14 |
Family
ID=44570582
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011101016376A Pending CN102184409A (en) | 2011-04-22 | 2011-04-22 | Machine-vision-based passenger flow statistics method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102184409A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663491A (en) * | 2012-03-13 | 2012-09-12 | 浙江工业大学 | Method for counting high density population based on SURF characteristic |
| CN103813138A (en) * | 2012-11-12 | 2014-05-21 | 安讯士有限公司 | Monitoring method and camera |
| CN104063253A (en) * | 2014-07-07 | 2014-09-24 | 无锡智广厦科技有限公司 | Method for automatic statistics of passenger flow, all-in-one machines and distributed system for automatic statistics of passenger flow |
| US9098769B2 (en) | 2012-03-21 | 2015-08-04 | Nec (China) Co., Ltd. | Method and a device for objects counting |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140623A (en) * | 2007-09-29 | 2008-03-12 | 华为技术有限公司 | Video frequency objects recognition method and system based on supporting vectors machine |
| CN101540892A (en) * | 2009-04-23 | 2009-09-23 | 上海中安电子信息科技有限公司 | Method for people counting in doorway on DSP video gathering device |
| CN101872422A (en) * | 2010-02-10 | 2010-10-27 | 杭州海康威视软件有限公司 | People flow rate statistical method and system capable of precisely identifying targets |
| CN201766663U (en) * | 2010-03-30 | 2011-03-16 | 苏州市职业大学 | Residential property monitoring system |
-
2011
- 2011-04-22 CN CN2011101016376A patent/CN102184409A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101140623A (en) * | 2007-09-29 | 2008-03-12 | 华为技术有限公司 | Video frequency objects recognition method and system based on supporting vectors machine |
| CN101540892A (en) * | 2009-04-23 | 2009-09-23 | 上海中安电子信息科技有限公司 | Method for people counting in doorway on DSP video gathering device |
| CN101872422A (en) * | 2010-02-10 | 2010-10-27 | 杭州海康威视软件有限公司 | People flow rate statistical method and system capable of precisely identifying targets |
| CN201766663U (en) * | 2010-03-30 | 2011-03-16 | 苏州市职业大学 | Residential property monitoring system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663491A (en) * | 2012-03-13 | 2012-09-12 | 浙江工业大学 | Method for counting high density population based on SURF characteristic |
| CN102663491B (en) * | 2012-03-13 | 2014-09-03 | 浙江工业大学 | Method for counting high density population based on SURF characteristic |
| US9098769B2 (en) | 2012-03-21 | 2015-08-04 | Nec (China) Co., Ltd. | Method and a device for objects counting |
| CN103813138A (en) * | 2012-11-12 | 2014-05-21 | 安讯士有限公司 | Monitoring method and camera |
| CN103813138B (en) * | 2012-11-12 | 2015-10-14 | 安讯士有限公司 | Supervision method and video camera |
| CN104063253A (en) * | 2014-07-07 | 2014-09-24 | 无锡智广厦科技有限公司 | Method for automatic statistics of passenger flow, all-in-one machines and distributed system for automatic statistics of passenger flow |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10909695B2 (en) | System and process for detecting, tracking and counting human objects of interest | |
| CN102054167B (en) | All-weather multipath channel pedestrian flow monitoring system based on wireless infrared monitoring | |
| CN107229894B (en) | Intelligent video monitoring method and system based on computer vision analysis technology | |
| CN102324024B (en) | Airport passenger recognition and positioning method and system based on target tracking technique | |
| Porikli et al. | Video surveillance: past, present, and now the future [DSP Forum] | |
| CN101751677B (en) | Target continuous tracking method based on multi-camera | |
| CN108764167A (en) | A kind of target of space time correlation recognition methods and system again | |
| CN101859436B (en) | Large-amplitude regular movement background intelligent analysis and control system | |
| CN102915638A (en) | Surveillance video-based intelligent parking lot management system | |
| CN101552910A (en) | Lave detection device based on comprehensive computer vision | |
| CN106127292B (en) | Flow method of counting and equipment | |
| CN100435160C (en) | Video image processing method and system for real-time sampling of traffic information | |
| CN202013603U (en) | Statistical device for passenger flow information | |
| CN106372570A (en) | Visitor flowrate statistic method | |
| CN113033293A (en) | Access & exit management equipment, system and method | |
| CN103646250A (en) | Pedestrian monitoring method and device based on distance image head and shoulder features | |
| Stahlschmidt et al. | Applications for a people detection and tracking algorithm using a time-of-flight camera | |
| CN102104771A (en) | Multi-channel people stream rate monitoring system based on wireless monitoring | |
| CN108710827A (en) | A kind of micro- police service inspection in community and information automatic analysis system and method | |
| US20200074157A1 (en) | Techniques for automatically identifying secondary objects in a stereo-optical counting system | |
| CN106652291A (en) | Indoor simple monitoring and alarming system and method based on Kinect | |
| CN102184409A (en) | Machine-vision-based passenger flow statistics method and system | |
| US20230046840A1 (en) | Vehicular access control based on virtual inductive loop | |
| KR100900494B1 (en) | System for movement tracing of moving object and service method thereof | |
| CN116245911A (en) | Video offline statistics method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20110914 |