CN105574499B - A kind of number detection statistics method and system based on SOC - Google Patents
A kind of number detection statistics method and system based on SOC Download PDFInfo
- Publication number
- CN105574499B CN105574499B CN201510934723.3A CN201510934723A CN105574499B CN 105574499 B CN105574499 B CN 105574499B CN 201510934723 A CN201510934723 A CN 201510934723A CN 105574499 B CN105574499 B CN 105574499B
- Authority
- CN
- China
- Prior art keywords
- image
- soc
- unit
- result
- people
- 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.)
- Expired - Fee Related
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06M—COUNTING MECHANISMS; COUNTING OF OBJECTS NOT OTHERWISE PROVIDED FOR
- G06M1/00—Design features of general application
- G06M1/27—Design features of general application for representing the result of count in the form of electric signals, e.g. by sensing markings on the counter drum
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/457—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by analysing connectivity, e.g. edge linking, connected component analysis or slices
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The number detection statistics method and system based on SOC that the present invention relates to a kind of.System includes image acquiring device, FPGA processing unit, microprocessor and unit as the result is shown;Described image acquisition device is input in the memory being connected with the FPGA unit for acquiring real time video data;The microprocessor control FPGA unit carries out number detection statistics method, detection, identification and tracking including moving target;The unit as the result is shown is for showing result.Efficiently real-time demographics can be achieved in the present invention.
Description
Technical field
The present invention relates to stream of people's detection technique field, more particularly to a kind of number detection statistics method based on SOC and
System.
Background technique
With the development and progress of modern society, the density of world population is increasing.Simultaneously with living standard
It increasingly improves, people's lives mode is also gradually changing.The consciousness of the chance that people go out, tourism gradually increases.For
This, the flow of the people on the ground such as public place such as market, scenic spot, railway station incrementally increases.And the flow of the people of these public places is excessive
When, it will cause very big security risk, it is crowded and the generation of accidents such as trample to easily cause the stream of people, causes serious safety
Accident.Because for the research of stream of people's behavior, detection to the stream of people, tracking and statistics increasingly cause people in modern society
Attention, also play increasingly consequence.
At the same time, the automation stream of people detection and analysis based on human motion analysis can largely save manpower and object
Power cost realizes continual automatic monitoring, statistics and alarm.By the analysis and prediction to human body behavior, the statistics of the stream of people,
Public place inlet and outlet number can be controlled in real time, to avoid and prevent the generation of peril.Meanwhile to people
The detection and analysis of stream can also realize otherwise application.Building designers are such as helped to optimize the structure design of building,
Pedestrian's monitoring of traffic intersection etc..Therefore to the detection of the stream of people, statistics, analysis for expanding economy and society stabilization,
The peril prevention and stream of people's control of especially public place have vital effect.
With the development of SOC technology, occur integrating the big rule of the microprocessors such as ARM Yu FPGA algorithm process unit
Mould embedded chip.The collaborative design of its software and hardware can either play the calculating advantage of FPGA in terms of processing, and can be abundant
Utilize advantage of ARM in terms of handling complex control algorithm, operation.This system is realized using hardware, it can be with
The real-time and processing speed of stream of people's statistics are greatly improved, realizes efficiently stream of people's statistics in real time.
Summary of the invention
The number detection statistics method and system based on SOC that technical problem to be solved by the invention is to provide a kind of, can
Realize efficiently real-time demographics.
The technical solution adopted by the present invention to solve the technical problems is: providing a kind of number detection statistics based on SOC
Method, comprising the following steps:
(1) video data is acquired by image acquiring device, and according to the installation site of image acquiring device, to determine back
Scape zone of action and goal activities region;
(2) Gaussian Background modeling is carried out to the image in background activity region, realizes the real-time update of background image;
(3) it is compared using background image with the current target image detected, prospect is obtained by background subtraction,
It is marked according to foreground picture of the background activity region to acquisition;
(4) binary conversion treatment is carried out to the foreground picture of acquisition, and connected component labeling is carried out to the binary image of acquisition;
(5) assume that K indicates k-th connected domain, N indicates the number of connected domain;Determine whether K is less than or equal to N, if K is small
In or equal to N, then follow the steps (6), it is no to then follow the steps (7);
(6) in connected domain K, local number P is calculatedK, in which: PK=Count (K)/Number;Count (K) is K
The pixel number of a connected domain;Number is the picture that the minimum circumscribed rectangle in the connected domain as unit of characteristics of human body includes
Vegetarian refreshments number;If PK>=1, by K plus 1, enter step (7);Otherwise, it is considered as calculating mistake, pedestrian target is not present in this connected domain;
(7) whole number P in a frame image is calculated, whereinThat is whole number is equal to the frame in a frame image
The calculated local number P of all connected domains in imageKSummation.
It further include poor to each frame progress in video according to background activity region between the step (1) and step (2)
The step of false retrieval is surveyed.
The step (2) further includes the installation site according to image acquiring device, is distorted and carries out to affine transformation bring
The step of amendment.
Using the real-time and concurrency of hardware, continuous three frames image is handled, and to the knot that three frame images obtain
Fruit carries out comprehensive analysis, realizes the tracking of crowd.
In the local number of calculating of the step (6) using the shoulder two sides of people relative to the symmetrical characteristic in head by image
In people effectively identify and come with object.
The technical solution adopted by the present invention to solve the technical problems is: also providing a kind of number detection system based on SOC
Meter systems, including image acquiring device, FPGA processing unit, microprocessor and unit as the result is shown;Described image acquisition device
For acquiring real time video data, and it is input in the memory being connected with the FPGA unit;The microprocessor control
FPGA unit carries out the processing such as the above-mentioned number detection statistics method based on SOC;The unit as the result is shown is for showing
As a result.
The unit as the result is shown is exported by WIFI module, and user takes by WEB server in conjunction with wechat subscription number
The channel for building user and terminal, by the way that the current stream of people can be inquired into public platform.
Beneficial effect
Due to the adoption of the above technical solution, compared with prior art, the present invention having the following advantages that and actively imitating
Fruit: it is this system compact, light, it is easily installed aloft (at the top of ceiling, pillar etc.), can effectively avoid the stream of people from blocking and ask
Topic.The present invention is during to collected video image analysis, using the real-time and concurrency of hardware to continuous three frames figure
Comprehensive analysis is carried out as being handled, and to the result that three frame images obtain, realizes that the tracking of crowd, reduction are obtained by a frame image
The error of the statistical result arrived, to more accurately obtain real-time demographics result.In addition, present invention incorporates wechat subscription
Number, build the new tunnel of user and terminal interaction.The wireless communication for realizing system and control centre, avoids wire transmission
The problems such as bring installation limitation.As long as user, which opens wechat, can be achieved with the use to system.
Detailed description of the invention
Fig. 1 is system structure diagram of the invention;
Fig. 2 is geometric distortion schematic diagram caused by the different position of system;
Fig. 3 is the minimum circumscribed rectangle definition figure as unit of characteristics of human body;
Fig. 4 is flow chart of the invention.
Specific embodiment
Present invention will be further explained below with reference to specific examples.It should be understood that these embodiments are merely to illustrate the present invention
Rather than it limits the scope of the invention.In addition, it should also be understood that, after reading the content taught by the present invention, those skilled in the art
Member can make various changes or modifications the present invention, and such equivalent forms equally fall within the application the appended claims and limited
Range.
It is as shown in Figure 1 system entire block diagram, including image acquiring device, FPGA processing unit, microprocessor and result
Display unit.Image acquiring device is input to memory using the infrared camera for being placed in eminence, acquisition real time video data;It is micro-
Picture frame is read in processor control, and FPGA carries out the processing of stream of people's detection algorithm.Stream of people's detection algorithm mainly includes moving target
Detection, identification and tracking, demographics result.Unit is exported by WIFI module as the result is shown, and user can be by WEB service
Device builds the channel of user and terminal in conjunction with wechat subscription number, by that can inquire current flow of the people into public platform.Work as people
When flowing through more, alarm function can also be provided.
It is illustrated in figure 2 region different location and obtains geometric distortion caused by image.It, need to be to affine change for different location
Bring distortion is changed to be modified.When background activity region is zone of action immediately below camera, captured image does not produce substantially
Raw distortion;In other positions, with height locating for system, the difference of angle and moving target present position is captured
Image can be distorted.Therefore when the sky between each respective pixel in the spatial relationship in original scene between each section and image
Between relationship it is inconsistent when, it is former between pixel to retrieve to need to correct the total each location of pixels of distorted image by geometric transformation
Carry out due spatial relationship.Can be according to the position of system, the different zones of target obtain different correction factors, to capture
Image is modified.
The minimum circumscribed rectangle definition figure being illustrated in figure 3 as unit of characteristics of human body.In connected domain analysis, the present invention
In view of the shoulder two sides of people are relative to the symmetrical characteristic in head, the people in image can be come with other object identifications (such as row
People hauls situations such as dragging case, can eliminate erroneous judgement).In connected domain K, local number P is calculatedK.Wherein: PK=Count (K)/
Number.Count (K) is the pixel number of k-th connected domain.Number be the connected domain in as unit of characteristics of human body
The pixel number that minimum circumscribed rectangle includes.If PK>=1, by K plus 1, calculate the number in next connected domain.Otherwise, depending on
To calculate mistake, pedestrian target is not present in this connected domain.
It is illustrated in figure 4 demographics implementation flow chart.After installing system, according to installation site, pass through corresponding four
Coordinate is put to mark and determine background activity region (unmanned zone of action) and goal activities region.According to the background activity of acquisition
Region can carry out error detection to each frame in video.Judge whether it is erroneous frame.If so, abandoning this frame.Erroneous frame is
There is situations such as blank screen, displacement, snowflake caused video frame in camera.Then Gaussian Background is carried out to the image of zone of action to build
Mould realizes the real-time update of background.According to the different location of system installation, the distortion of affine transformation bring is modified.This
Place can refer to attached drawing 2.Background image is compared with the current target image detected, before being obtained by background subtraction
Scape.Foreground picture is marked according to the background activity region of acquisition, 0 is labeled as to the pixel value of unmanned zone of action.To preceding
Scape figure carries out binary conversion treatment, and obtained background is black, artificial white.And connected component labeling is carried out to binary image.It is false
If K indicates k-th connected domain, N indicates the number of connected domain.Determine whether K is less than or equal to N.Start to calculate connection if setting up
Local number in the K of domain.In connected domain analysis, the present invention in view of the shoulder two sides of people are relative to the symmetrical characteristic in head,
People in image can be come with other object identifications.In connected domain K, local number P is calculatedK.Wherein: PK=Count (K)/
Number.Count (K) is the pixel number of k-th connected domain.Number be the connected domain in as unit of characteristics of human body
The pixel number that minimum circumscribed rectangle includes.It can refer to attached drawing 3.If PK>=1, by K plus 1.Otherwise, it is considered as calculating mistake, this
Pedestrian target is not present in connected domain.Finally calculate whole number P in a frame image.That is whole man in a frame image
Number is equal to the calculated local number P of connected domain all in the frame imageKSummation.The present invention is to collected video figure
As being handled using the real-time and concurrency of hardware continuous three frames image, and obtained to three frame images in analytic process
Result carry out comprehensive analysis, the tracking of crowd is realized, to more accurately obtain real-time demographics result.
Result is exported real-time statistics result by WIFI module by the present invention.By WEB server, subscribed in conjunction with wechat
Number, the channel of user and terminal is built, the wireless communication of system and control centre is realized.User is by entering public platform
Inquire current flow of the people.When the stream of people is excessive, alarm function can also be provided.
In conjunction with wechat, realizing wireless communication, specific step is as follows:
1: personally opening a wechat subscription number, user can be with SOC demographics after paying close attention to the subscription number
The WEB server interaction of system.
2: after user pays close attention to subscription number, wechat platform can be sent commands to by sending message in subscription number.
3: platform forwards the message to the message interface built on SOC passenger number statistical system, interface and parsing information.
Judge that the type and content of message execute the script finished writing in advance.
4: after script execution, returning to message.Message is encapsulated as XML format by platform interface, then message
Wechat platform is transmitted to according to XML format.It will be replied message by wechat platform and be sent to user.
It is not difficult to find that the present invention is during to collected video image analysis, the real-time of hardware and parallel is utilized
Property continuous three frames image is handled, and three frame images are obtained result carry out comprehensive analysis, realize the tracking of crowd, drop
The error of the low statistical result obtained by a frame image, to more accurately obtain real-time demographics result.In addition, of the invention
Wechat subscription number is combined, the new tunnel of user and terminal interaction has been built.The wireless communication of system and control centre is realized,
The problems such as avoiding the installation limitation of wire transmission bring.As long as user, which opens wechat, can be achieved with the use to system.
Claims (4)
1. a kind of number detection statistics method based on SOC, which comprises the following steps:
(1) video data is acquired by image acquiring device, and according to the installation site of image acquiring device, to determine background living
Dynamic region and goal activities region;
(2) Gaussian Background modeling is carried out to the image in background activity region, realizes the real-time update of background image, and according to image
The installation site of acquisition device is modified the distortion of affine transformation bring;
(3) it is compared using background image with the current target image detected, prospect is obtained by background subtraction, according to
The foreground picture of acquisition is marked in background activity region;
(4) binary conversion treatment is carried out to the foreground picture of acquisition, and connected component labeling is carried out to the binary image of acquisition;
(5) assume that K indicates k-th connected domain, N indicates the number of connected domain;Determine K whether be less than or equal to N, if K be less than or
Equal to N, then follow the steps (6), it is no to then follow the steps (7);
(6) in connected domain K, local number P is calculatedK, in which: PK=Count (K)/Number;Count (K) is k-th company
The pixel number in logical domain;Number is the pixel that the minimum circumscribed rectangle in the connected domain as unit of characteristics of human body includes
Number;If PK>=1, by K plus 1, enter step (7);Otherwise, it is considered as calculating mistake, pedestrian target is not present in this connected domain;Its
In, it calculates and is effectively known the people in image with object relative to the symmetrical characteristic in head using the shoulder two sides of people in local number
It does not come;
(7) whole number P in a frame image is calculated, whereinThat is whole number is equal to the frame image in a frame image
In the calculated local number P of all connected domainsKSummation;
(8) real-time and concurrency for utilizing hardware, are handled continuous three frames image, and to the result that three frame images obtain
Comprehensive analysis is carried out, realizes the tracking of crowd.
2. the number detection statistics method according to claim 1 based on SOC, which is characterized in that the step (1) and step
Suddenly further include the steps that carrying out error detection to each frame in video according to background activity region between (2).
3. a kind of number detection statistics system based on SOC, which is characterized in that including image acquiring device, FPGA processing unit,
Microprocessor and as the result is shown unit;Described image acquisition device for acquiring real time video data, and be input to it is described
In the connected memory of FPGA unit;The microprocessor control FPGA unit is carried out as described in any in claim 1-2
The processing of number detection statistics method based on SOC;The unit as the result is shown is for showing result.
4. the number detection statistics system according to claim 3 based on SOC, which is characterized in that described single as the result is shown
Member is exported by WIFI module, and user builds the channel of user and terminal, pass through by WEB server in conjunction with wechat subscription number
The current stream of people can be inquired into public platform.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510934723.3A CN105574499B (en) | 2015-12-15 | 2015-12-15 | A kind of number detection statistics method and system based on SOC |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510934723.3A CN105574499B (en) | 2015-12-15 | 2015-12-15 | A kind of number detection statistics method and system based on SOC |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105574499A CN105574499A (en) | 2016-05-11 |
CN105574499B true CN105574499B (en) | 2019-08-20 |
Family
ID=55884607
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510934723.3A Expired - Fee Related CN105574499B (en) | 2015-12-15 | 2015-12-15 | A kind of number detection statistics method and system based on SOC |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105574499B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107451678A (en) * | 2017-07-03 | 2017-12-08 | 温州市正立建筑设计有限公司 | A kind of method judged for architectural design prioritization scheme |
CN110838134B (en) * | 2019-10-10 | 2020-09-29 | 北京海益同展信息科技有限公司 | Target object statistical method and device, computer equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101303727A (en) * | 2008-07-08 | 2008-11-12 | 北京中星微电子有限公司 | Intelligent management method based on video human number Stat. and system thereof |
CN102289805A (en) * | 2011-06-10 | 2011-12-21 | 北京云加速信息技术有限公司 | Video-based method for detecting crowd density in subway |
CN102542289A (en) * | 2011-12-16 | 2012-07-04 | 重庆邮电大学 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
CN104021599A (en) * | 2014-05-19 | 2014-09-03 | 沈阳工业大学 | Classroom attendance management system based on campus WiFi and people counting method |
CN104813339A (en) * | 2012-09-12 | 2015-07-29 | 威智伦富智堡公司 | Methods, devices and systems for detecting objects in a video |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013033317A (en) * | 2011-08-01 | 2013-02-14 | Sony Corp | Image processing apparatus, image processing method, and program |
-
2015
- 2015-12-15 CN CN201510934723.3A patent/CN105574499B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101303727A (en) * | 2008-07-08 | 2008-11-12 | 北京中星微电子有限公司 | Intelligent management method based on video human number Stat. and system thereof |
CN102289805A (en) * | 2011-06-10 | 2011-12-21 | 北京云加速信息技术有限公司 | Video-based method for detecting crowd density in subway |
CN102542289A (en) * | 2011-12-16 | 2012-07-04 | 重庆邮电大学 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
CN104813339A (en) * | 2012-09-12 | 2015-07-29 | 威智伦富智堡公司 | Methods, devices and systems for detecting objects in a video |
CN104021599A (en) * | 2014-05-19 | 2014-09-03 | 沈阳工业大学 | Classroom attendance management system based on campus WiFi and people counting method |
Non-Patent Citations (1)
Title |
---|
基于机器视觉的人数统计设计;袁孙正;《中国优秀硕士学位论文全文数据库 信息科技辑》;20111015(第10期);I138-792 |
Also Published As
Publication number | Publication date |
---|---|
CN105574499A (en) | 2016-05-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11182598B2 (en) | Smart area monitoring with artificial intelligence | |
US12008794B2 (en) | Systems and methods for intelligent video surveillance | |
CN103069434B (en) | For the method and system of multi-mode video case index | |
KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
CN101458871B (en) | Intelligent traffic analysis system and application system thereof | |
WO2019129255A1 (en) | Target tracking method and device | |
WO2016004673A1 (en) | Intelligent target recognition device, system and method based on cloud service | |
WO2021031954A1 (en) | Object quantity determination method and apparatus, and storage medium and electronic device | |
SG191954A1 (en) | An integrated intelligent server based system and method/systems adapted to facilitate fail-safe integration and /or optimized utilization of various sensory inputs | |
CN105611244A (en) | Method for detecting airport foreign object debris based on monitoring video of dome camera | |
US12094252B2 (en) | Occlusion-aware prediction of human behavior | |
CN102426785A (en) | Traffic flow information perception method based on contour and local characteristic point and system thereof | |
CN111967396A (en) | Processing method, device and equipment for obstacle detection and storage medium | |
CN104159088A (en) | System and method of remote monitoring of intelligent vehicle | |
CN110795975B (en) | Face false detection optimization method and device | |
Yaghoobi Ershadi et al. | Vehicle tracking and counting system in dusty weather with vibrating camera conditions | |
CN114429596A (en) | Traffic statistical method and device, electronic equipment and storage medium | |
Kostić et al. | Smart city intersections: Intelligence nodes for future metropolises | |
CN105574499B (en) | A kind of number detection statistics method and system based on SOC | |
Li et al. | Intelligent transportation video tracking technology based on computer and image processing technology | |
CN114627409A (en) | Method and device for detecting abnormal lane change of vehicle | |
Mekni et al. | Real-time smart parking integration in intelligent transportation systems (its) | |
KR102240638B1 (en) | Parking guidance method and system using boundary pixel data estimated in vehicle image and analysis of vehicle model viewpoint | |
CN112422895A (en) | Image analysis tracking and positioning system and method based on unmanned aerial vehicle | |
CN116611977A (en) | Method and device for tracing and rescuing path of disjunctor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190820 Termination date: 20211215 |
|
CF01 | Termination of patent right due to non-payment of annual fee |