CN102542289A - Pedestrian volume statistical method based on plurality of Gaussian counting models - Google Patents
Pedestrian volume statistical method based on plurality of Gaussian counting models Download PDFInfo
- Publication number
- CN102542289A CN102542289A CN2011104233492A CN201110423349A CN102542289A CN 102542289 A CN102542289 A CN 102542289A CN 2011104233492 A CN2011104233492 A CN 2011104233492A CN 201110423349 A CN201110423349 A CN 201110423349A CN 102542289 A CN102542289 A CN 102542289A
- Authority
- CN
- China
- Prior art keywords
- moving target
- digital
- image
- analogue type
- gaussmeter
- 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.)
- Granted
Links
Images
Abstract
The invention relates to intelligent video surveillance and image processing and analysis, discloses a pedestrian volume statistical method, and comprises establishing a plurality of Gaussian counting models by utilizing training video sequence image samples with people number marks and performing real-time pedestrian volume statistics on videos with unknown people numbers based on the plurality of Gaussian counting models. The pedestrian volume statistical method particularly comprises the steps of firstly extracting a prospect moving target according to moving target detection, extracting eigenvectors according to moving target area and characteristics including lengths and widths of an external rectangular frame, then establishing the plurality of Gaussian counting models based on an eigenvector set, and finally analyzing numbers of pedestrians contained in an unknown moving target area based on the plurality of Gaussian counting models to achieve pedestrian volume statistics. By establishing the plurality of Gaussian counting models, the pedestrian volume statistical method avoids difficulties caused by identification and tracking of singe pedestrian, can perform statistics of the numbers of the pedestrians contained in moving target areas in different detection areas well, improves statistical accuracy of the numbers of the pedestrians, and then improves accuracy of the pedestrian volume statistics.
Description
Technical field
The present invention relates to technical field of image processing, relate in particular to a kind of method and system of people flow rate statistical.
Background technology
In the existing video monitoring, the just simple realization video transmission of great majority relies on eye-observation to realize work such as scene monitoring and counting then.There is a large amount of weak points in this manpower monitor mode, and as more uninteresting, the monitor staff is easy to produce tired and causes work mistake, and in addition, along with human cost improves, relying on manpower to monitor these means of counting will be no longer suitable.
The present method that is adopted based on the people flow rate statistical system of computer vision can be divided three classes: one is based on the method that pedestrian detection is followed the tracks of; Two are based on the method for unique point trajectory clustering; Three are based on the method that low-level feature returns.
The core of the method for following the tracks of based on pedestrian detection is multi-target detection; This method is that the method through background difference or machine learning obtains foreground area, adopts motion morphology to unite then to cut apart or the method for template matches is accomplished the task of people flow rate statistical.Generally, this algorithm can obtain the higher detection precision.Be the method for being mentioned in 201010114826.2 the one Chinese patent application file like application number, at first present image carried out number of people rough detection, then the rough detection result is carried out the edge feature fine screening and handle through sorter; Though can effectively improve the verification and measurement ratio of the number of people through said method,, when crowd density higher; When situation such as blocking; Have number of people omission, perhaps the situation of many inspections finally causes testing result not accurate enough; And this method calculated amount is bigger, is difficult to real-time processing.
At first pass through to follow the tracks of some unique points based on the method for unique point trajectory clustering, the unique point track with consistent kinetic characteristic is carried out the purpose that cluster analysis reaches demographics.This algorithm can effectively reduce the influence at video camera visual angle.Yet unique point itself is difficult to reliable and stable tracking, so this algorithm statistical precision is lower.
Method based on low-level feature returns at first utilizes the background subtraction branch to obtain foreground area; Calculate characteristic such as area in the foreground area, edge, texture etc. then, set up the funtcional relationship of characteristic and flow of the people at last through various regression functions like linearity, Gaussian process recurrence, neural network etc.This algorithm has been skipped the detection tracing process for single pedestrian's target, has reduced computation complexity, can reach the real-time requirement to a certain extent.But its versatility is not ideal enough, and the dependence of statistical accuracy and foreground pixel extraction is bigger, so this method is difficult to obtain accurate number information.
In sum, difficult point mainly is how the mobile crowd with greater density to be carried out the statistics of degree of precision in the people flow rate statistical method of the prior art, and the complexity of algorithm can not be too high, and satisfies real-time application demand.
Summary of the invention
The present invention is directed to the problems referred to above that exist in the existing people flow rate statistical technology based on computer vision; A kind of real-time people flow rate statistical method based on many gaussmeters digital-to-analogue type is proposed, to solve existing people flow rate statistical scheme to the mobile not statistical uncertainty true problem of crowd of higher density.
The present invention solves the problems of the technologies described above and adopts following technical scheme:
A kind of people flow rate statistical method based on many gaussmeters digital-to-analogue type comprises: input picture pre-service, moving object detection, moving target proper vector are extracted and steps such as the foundation of many gaussmeters digital-to-analogue type, motion target tracking and people flow rate statistical.Be specially:
In actual detected, often only interested in certain zone in the scene, (region of interest, ROI), follow-up all images is handled operation and all in this region-of-interest, is accomplished therefore at first to choose region-of-interest.Region-of-interest is divided into the equal-sized detection subregion of a plurality of areas, adopts foreground image and background image to make the moving target detecting method of difference, obtain the foreground moving target; The moving target that will be under the jurisdiction of a connected domain comes out with the rectangle frame mark; Extract this rectangle frame moving target proper vector and obtain set of eigenvectors; Extract the moving target proper vector, the target feature vector composition characteristic vector set that has identical number in the same subregion; Set up corresponding gaussmeter digital-to-analogue type based on set of eigenvectors, the gaussmeter digital-to-analogue type that on same subregion, obtains is formed the Gauss model subclass, and the model subclass on all subregions is formed final many gaussmeters digital-to-analogue type; During people flow rate statistical, detection line is set, the sequence of video images of unknown number is carried out image pre-service and moving object detection; To carrying out target following, judge whether moving target boundary rectangle frame intersects with detection line, if non-intersect with the crossing motion target area of detection line; Then the next frame image is handled; Till its boundary rectangle frame arrives detection line,, extract the proper vector of current moving target if intersect; In each two field picture of tracing process, extract the proper vector of current moving target; According to the residing subregion of current moving target; Adopt corresponding Gauss model subclass to analyze the number in the current moving target; Adopt fast target to follow the tracks of correlating method and obtain the corresponding count queue in this target area, and deposit the number that obtains the corresponding formation of in this target area; When moving target leaves detection line, calculate the mean value of number in the formation, obtain people flow rate statistical.
Region-of-interest in the scene is divided into a series of detection subregions; As may be partitioned into
detection subregion that individual area is identical; Adopt
and represent
OK; The subregion (
of
row;
); Wherein
;
is according to the size of field of detection; And pedestrian's size in the visual field, confirm its value.Generally speaking, detect subregion area size and can be three to four pedestrian's area sizes in the field of detection.Video image after cutting apart is carried out The disposal of gentle filter, reduce The noise.
Extract background image, current frame image and background image in the detection subregion are carried out difference processing, obtain difference image.
Calculate in the difference image standard variance of pixel value between different gray areas, and confirm segmentation threshold, image is cut apart and bianry image morphology is handled, extract the foreground moving target according to the maximal value in these standard variances.
The moving target that will be under the jurisdiction of a connected domain is remembered out with its boundary rectangle collimation mark, and is extracted the proper vector
of this motion target area.
According to the residing position of rectangle frame central point; The detection subregion of confirming the inferior appearance of moving target
place is for
; Proper vector
when pedestrian's number is
(comprises the moving target area and is
(number of pixels) in the embodiment of the invention; Rectangle frame long
, wide
three characteristics), the proper vector composition characteristic vector set
that will have identical pedestrian's number.Set up corresponding gaussmeter digital-to-analogue type
based on set of eigenvectors.Be specially:
Adopt formula:
(wherein
expression detection subregion is
; When the detection number is
; Corresponding sample proper vector number); The mean vector
of calculated characteristics vector set
; Adopt formula according to mean vector:
calculates its covariance matrix
; According to formula:
; Setting up and detecting subregion is
; When detection pedestrian number was
, corresponding gaussmeter digital-to-analogue type was
.
The gaussmeter digital-to-analogue type
of corresponding same detection subregion is formed gaussmeter digital-to-analogue type subclass
, a plurality of gaussmeter digital-to-analogue type subclass
are formed many gaussmeters digital-to-analogue type
.
When carrying out people flow rate statistical; Detection line is set; Sequence of video images to unknown number carries out image pre-service and moving object detection; Judge whether moving target boundary rectangle frame intersects with detection line; If it is non-intersect; Then the next frame image is handled, till its boundary rectangle frame arrives detection line, if intersect; Then extract the proper vector of current moving target; According to its residing subregion
, call corresponding gaussmeter digital-to-analogue type subclass
and calculate the pedestrian's number that comprises in the current moving target, and deposit the number that obtains in this moving target corresponding queues; Continue the pursuit movement target, in subsequent frame, call corresponding gaussmeter digital-to-analogue type subclass, calculate pedestrian's number that this moving target comprises and deposit corresponding queues in according to its residing subregion; When moving target leaves detection line, carry out people flow rate statistical, that is, calculate the average of number in the formation, obtain the number of current tracking target.
The present invention sets up gaussmeter digital-to-analogue type and carries out the calculating of moving target number through to the moving object detection in the region-of-interest, can when the flow of the people of big density, carry out the people flow rate statistical of degree of precision, and reduce computation complexity, reaches the real-time requirement.
Description of drawings
In Fig. 1 embodiment of the invention based on the process flow diagram of the people flow rate statistical method of many gaussmeters digital-to-analogue type;
Many gaussmeters digital-to-analogue type establishment stage process flow diagram in Fig. 2 embodiment of the invention;
Foreground moving target detection process flow diagram in Fig. 3 embodiment of the invention;
Target following synoptic diagram in Fig. 4 embodiment of the invention.
Embodiment
Below in conjunction with accompanying drawing and specific embodiment the present invention is described in detail.
A kind of people flow rate statistical based on many gaussmeters digital-to-analogue type that the present invention proposes is used for real-time monitoring system.Obtain the foreground moving target through motion detection,, realize people flow rate statistical according to the number in many gaussmeters digital-to-analogue type analysis present image.
Fig. 1 is based on the people flow rate statistical process flow diagram of many gaussmeters digital-to-analogue type in the embodiment of the invention.As shown in Figure 1, before carrying out people flow rate statistical, need set up many gaussmeters digital-to-analogue type in the embodiment of the invention.
Region-of-interest in the scene is divided into a series of detection subregions.Video image after cutting apart is carried out The disposal of gentle filter.
Extract background image, current frame image and background image in the detection subregion are carried out difference processing, obtain difference image.
Calculate in the difference image standard variance of pixel value between different gray areas, and confirm segmentation threshold, image is cut apart and bianry image morphology is handled, extract the foreground moving target according to the maximal value in these standard variances.
The moving target that will be under the jurisdiction of a connected domain is remembered out with its boundary rectangle collimation mark, and is extracted the proper vector
of this motion target area.
According to the residing position of rectangle frame central point; The detection subregion of confirming the inferior appearance of moving target
place is for
; (proper vector comprises the moving target area and is
(number of pixels) proper vector
of pedestrian's number when
in the embodiment of the invention; Rectangle frame long
, wide
three characteristics); To have the proper vector composition characteristic vector set
of identical pedestrian's number, set up corresponding gaussmeter digital-to-analogue type
based on set of eigenvectors.Be specially:
Adopt formula:
(wherein
expression detection subregion is
; When the detection number is
; Corresponding sample proper vector number); The mean vector
of calculated characteristics vector set
; Adopt formula according to mean vector:
calculates its covariance matrix
; According to formula:
; Setting up and detecting subregion is
; When detection pedestrian number was
, corresponding gaussmeter digital-to-analogue type was
.
The gaussmeter digital-to-analogue type
of corresponding same detection subregion is formed gaussmeter digital-to-analogue type subclass
, a plurality of gaussmeter digital-to-analogue type subclass
are formed many gaussmeters digital-to-analogue type
.
When carrying out people flow rate statistical; Detection line is set; Sequence of video images to unknown number carries out image pre-service and moving object detection; Judge whether moving target boundary rectangle frame intersects with detection line; If it is non-intersect; Then the next frame image is handled, till its boundary rectangle frame arrives detection line, if intersect; Then extract the proper vector of current moving target; According to its residing subregion
, call corresponding gaussmeter digital-to-analogue type subclass
and calculate the pedestrian's number that comprises in the current moving target, and deposit the number that obtains in this moving target corresponding queues; Continue the pursuit movement target, in subsequent frame, call corresponding gaussmeter digital-to-analogue type subclass, calculate pedestrian's number that this moving target comprises and deposit corresponding queues in according to its residing subregion; When moving target leaves detection line, carry out people flow rate statistical, that is, calculate the average of number in the formation, obtain the number of current tracking target.
Fig. 2 sets up process flow diagram for many gaussmeters digital-to-analogue type in the embodiment of the invention, specifically comprises following several steps:
Obtain the sequence of video images of band number mark through shooting.
Surveyed area is divided.Select the region-of-interest of the video image of input; Then region-of-interest is divided into a series of detection subregions; Region-of-interest in the scene is divided into a series of detection subregions; As may be partitioned into
detection subregion that individual area is identical; Adopt
and represent
OK; The subregion (
of
row;
); Wherein
;
is according to the size of field of detection; And pedestrian's size in the visual field, confirm its value.Generally speaking, detect subregion area size and can be three to four pedestrian's area sizes in the field of detection.Video image after cutting apart is carried out The disposal of gentle filter, reduce The noise.
The image pre-service.
The input video sequence image of region-of-interest is converted into grayscale image sequence, and pre-processing module is carried out filtering to grayscale image sequence, the noise in the removal of images.The embodiment of the invention adopts Gauss smoothly image to be carried out filtering.
Moving object detection.Extract background image:, select suitable background image to extract according to the characteristics of monitoring scene.Current frame image and background image are carried out calculus of differences acquisition difference image, carry out carrying out image threshold segmentation and bianry image morphology and handle, obtain the foreground moving target.
Proper vector is extracted.
Obtain moving target area in the rectangle frame, and the length of rectangle frame and width.With the moving target area is (
); Rectangle frame long (
), wide (
) three latent structure proper vectors, i.e.
.The subregion at the moving target place of
expression
inferior appearance is
, the proper vector when pedestrian's number is
;
The same number of sub-regions within the same target connected domain (ie, having the same
) eigenvectors
integral feature vector set
.
Analyze each proper vector and concentrate vectorial number; Like vectorial number less than
; Then continue to extract proper vector; Wherein the value of
is the bigger the better, and rule of thumb generally gets
.When proper vector concentrates vectorial number more than or equal to
; According to set of eigenvectors, set up single gaussmeter digital-to-analogue type
.Single gaussmeter digital-to-analogue type
to set up process following:
Call formula:
mean vector
of calculated characteristics vector set
; Utilize formula according to mean vector:
calculate its covariance matrix
; Set up surveyed area for
according to covariance matrix; When detection pedestrian number is
; Corresponding gaussmeter digital-to-analogue type
:
; In the formula;
is random vector,
represent transposition.
After all vector sets
are set up corresponding gaussmeter digital-to-analogue type; The gaussmeter digital-to-analogue type
that is in same subregion (promptly having identical
) is formed gaussmeter digital-to-analogue type subclass
; All gaussmeter digital-to-analogue type subclass
are formed many gaussmeters digital-to-analogue type
, i.e.
.
After accomplishing the foundation of many gaussmeters digital-to-analogue type, can count model according to this and carry out people flow rate statistical.As shown in Figure 1, specifically comprise the steps:
Obtain video sequence image through single ccd imaging sensor through vertical the shooting.Surveyed area is carried out area dividing.
Detection line is set, preferably selects the center line of region-of-interest, like the line segment L1 of Fig. 4.Image is carried out pre-service, moving object detection.
If do not detect moving target, then go to next frame and continue to handle; If detect moving target; Judge then whether moving target boundary rectangle frame intersects with detection line; If it is non-intersect; Then go to next frame and continue processing, otherwise extract the proper vector
(the same) of current moving target with the proper vector of many gaussmeters digital-to-analogue type establishment stage.Estimate the number that current motion target area comprises, specifically comprise the steps:
Confirm its residing subregion
according to moving target boundary rectangle frame center point coordinate; In many gaussmeters digital-to-analogue type, obtain the corresponding gaussmeter digital-to-analogue type subclass
of subregion
; Promptly
is
;
;
, the set that
forms.Bring current proper vector
into model (Gaussian function)
respectively;
;
; Calculate in
; Then result of calculation is that the corresponding number of peaked model
is the number that current motion target area comprises, and its value equals
.That is:
Calculate the number
that comprises in the current moving target.Wherein, the maximal value in
representation model subclass
.
After the number that the acquisition motion target area comprises, can obtain the corresponding count queue in this target area through fast tracking method, and estimate that current acquisition number deposit in this count queue based on detection line.Statistical number of person is recorded in this detection target corresponding queues.
Be illustrated in figure 3 as foreground moving target detection process flow diagram, specifically comprise the steps:
Extract background image:, select suitable background image method for distilling according to the characteristics of monitoring scene.As: do not change or change very little particular surroundings (like the part indoor environment) for some backgrounds, directly the shooting background image then keeps background image constant.For the tangible scene of change of background (like natural scene); Use obtains background image based on histogrammic background modeling method; Be specially: the grey level histogram of statistical series image pixel value on same pixel position, the maximum gray-scale value of occurrence number is as the background pixel value of this point.
Obtain difference image: in the moving object detection processing module, carry out calculus of differences to current image frame of from sequence of video images, obtaining and background image, obtain difference image.
The difference image Threshold Segmentation: the difference image that obtains is 8 gray level images of standard; Therefore its pixel value scope is [0; 255]; Ask for successively that tonal range is respectively [0 in the difference image; 255], [1; 255] ..., [254; 255]; The standard variance of interior pixel value between 254 gray areas altogether; Calculate the maximal value
of these 254 standard variances then; And two threshold values
are set;
(
); (
,
can choose according to the inputted video image quality; Generally speaking;
;
); According to
,
confirm in the middle of threshold value T, ask for final segmentation threshold
through following formula;
,
Where
,
strike two respectively the maximum and minimum values.With
as threshold value; Image is cut apart, obtained bianry image.
Bianry image morphology processing: in bianry image, carry out morphologic filtering, carry out the zone merging to the target area of some fractures promptly through the less false target zone of some areas of morphological erosion operation deletion, and through expansive working.Then, adopt each connected region of 8 neighbour's connected domain searching algorithms search, carry out mark, obtain the foreground moving target thus with the boundary rectangle of connected domain.
The target fast tracking method is as shown in Figure 4; Comprise the steps: that specifically the purpose of target following is exactly to realize the association of motion target area in the two continuous frames image in the present embodiment, whether the moving target that intersects with detection line in two two field pictures before and after promptly judging is same target.
Showed under two nearer target conditions of single moving target and longitudinal separation the state in two two field pictures of front and back among Fig. 4.Among the figure, region R 1 is a video-input image; Region R 2 is region-of-interest (can select the region-of-interest of suitable shapes and size according to actual needs); Region R 3 is the boundary rectangle frame of moving target; Line segment L1 is a detection line, generally selects the center line of region-of-interest; Line segment L2 is the intersection section of moving target and detection line; P1, P2 are the summit in the upper right corner of external rectangle frame in front and back two frames (frame and
frame) image.
The coordinate of supposing P1 and P2 is respectively
and
(image coordinate system; Be that the image upper left corner is initial point; X axle positive dirction be level to the right; Y axle positive dirction is for vertically downward); Whether the moving target before and after differentiating through two constraint conditions in two two field pictures is same target, that is: the intersection section of motion target area in two two field pictures and detection line is overlapping or overlap before and after (1); (2) coordinate
.If the moving target in two two field pictures of front and back satisfies (1) and (2) two constraint conditions simultaneously, then be judged as same target, the person is not judged as different targets.For situation about only comprising, can realize correct judgement with constraint (1), shown in Fig. 4 (a) when a moving target.But,, must could realize correct judgement through constraint (2) although the intersection section of front and back two frames is overlapping to the situation of two nearer moving targets of the longitudinal separation of Fig. 4 (b).Because it is higher to handle frame frequency, (1) and (2) two constraint Rule of judgment can realize counting extremely greatly the correct judgement of situation, and this algorithm is simple, efficiently.
If moving target leaves detection line; Be that continuous three frames of moving target and detection line are when non-intersect; Satisfy the counting condition; Calculate the average
of pedestrian's number in the corresponding count queue of current moving target, obtain the number that current moving target comprises.The average of each formation pedestrian number that adds up then; Obtain total flow of the people
, thereby realize people flow rate statistical.
The above only is a preferred implementation of the present invention; Be noted that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; Can also make some improvement and retouching, these improvement and retouching also should be regarded as protection scope of the present invention.
Claims (7)
1. the people flow rate statistical method based on many gaussmeters digital-to-analogue type is characterized in that, comprises the steps, video image is divided into the detection subregion, and current frame image and background image are done difference, obtains difference image, obtains the foreground moving target; The moving target that will be under the jurisdiction of a connected domain comes out with the rectangle frame mark, extracts this rectangle frame moving target proper vector and obtains set of eigenvectors; Set up corresponding gaussmeter digital-to-analogue type based on set of eigenvectors
; The gaussmeter digital-to-analogue type that belongs to identical subregion is formed gaussmeter digital-to-analogue type subclass, and all gaussmeter digital-to-analogue type subclass constitute many gaussmeters digital-to-analogue type; Judge that whether moving target boundary rectangle frame intersects with detection line, if non-intersect, then handles the next frame image, till its boundary rectangle frame arrives detection line; If intersect; Extract the proper vector of current moving target,, analyze current moving target number with corresponding gaussmeter digital-to-analogue type subclass according to the residing subregion of current moving target; Adopt fast target to follow the tracks of correlating method and obtain the corresponding count queue in this target area; And deposit current moving target number in this formation, when target was left detection line, the mean value that calculates formation obtained the number that this target area comprises.
2. people flow rate statistical method according to claim 1; It is characterized in that; The foreground moving target of said acquisition is specially: extract background image, calculate in the difference image that detects present image and background image in the subregion maximal value of the standard variance of pixel value between each gray area, and confirm segmentation threshold with this; According to segmentation threshold difference image is cut apart, and bianry image morphology is handled acquisition foreground moving target.
3. people flow rate statistical method according to claim 1; It is characterized in that; Confirm that according to rectangle frame area, length and width inferior the appearing at of moving target
detect that subregion is
; Proper vector when pedestrian's number is
; The proper vector composition characteristic vector set
that will have identical pedestrian's number; Set up corresponding gaussmeter digital-to-analogue type
based on set of eigenvectors, and be classified as a gaussmeter digital-to-analogue type subclass
to
that be in same subregion.
4. people flow rate statistical method according to claim 1; It is characterized in that; The difference image Threshold Segmentation is specially: the gray level image that obtains difference image; Ask in the difference image standard variance of interior pixel value between gray area successively; The maximal value of standard variance
altogether; According to formula:
;
asks for segmentation threshold
; With
as threshold value; Image is cut apart; Obtain bianry image; Wherein,
,
threshold value (
) for being provided with.
5. according to claim 1 or 3 described flow statistical methods; It is characterized in that; Set up corresponding gaussmeter digital-to-analogue type
based on set of eigenvectors
; Be specially: call formula:
according to proper vector the mean vector
of calculated characteristics vector set
; Utilize formula according to mean vector:
calculating covariance matrix
; According to formula:
sets up that surveyed area is
; When detection pedestrian number was
, corresponding gaussmeter digital-to-analogue type was
.
6. according to one of them described people flow rate statistical method of claim 1-3; It is characterized in that; Analyze the number of current moving target with corresponding gaussmeter digital-to-analogue type subclass; Be specially: confirm residing subregion
according to moving target boundary rectangle frame center point coordinate; In many gaussmeters digital-to-analogue type, obtain the corresponding gaussmeter digital-to-analogue type subclass
of subregion
; I.e.
=(
;
;
;
), utilize formula:
7. people flow rate statistical method according to claim 1; It is characterized in that; Fast target is followed the tracks of correlating method; Be specially: suppose that P1 and P2 are respectively the summit, the upper right corner of moving target boundary rectangle in two two field pictures of front and back; Its coordinate is respectively
and
, as satisfies condition: the motion target area before and after (1) in two two field pictures and the intersection section of detection line are overlapping or overlap; (2) coordinate
then is same target.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110423349.2A CN102542289B (en) | 2011-12-16 | 2011-12-16 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110423349.2A CN102542289B (en) | 2011-12-16 | 2011-12-16 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102542289A true CN102542289A (en) | 2012-07-04 |
CN102542289B CN102542289B (en) | 2014-06-04 |
Family
ID=46349139
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201110423349.2A Active CN102542289B (en) | 2011-12-16 | 2011-12-16 | Pedestrian volume statistical method based on plurality of Gaussian counting models |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102542289B (en) |
Cited By (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103186903A (en) * | 2013-04-22 | 2013-07-03 | 苏州科技学院 | Moving object detecting and tracking system based on PID (Proportion Integration Differentiation) |
CN103400113A (en) * | 2013-07-10 | 2013-11-20 | 重庆大学 | Method for detecting pedestrian on expressway or in tunnel based on image processing |
CN103839035A (en) * | 2012-11-22 | 2014-06-04 | 富士通株式会社 | Person number statistical method and person number statistical system |
CN104123569A (en) * | 2014-07-21 | 2014-10-29 | 中国人民解放军63908部队 | Video person number information statistics method based on supervised learning |
CN104239905A (en) * | 2013-06-17 | 2014-12-24 | 上海盖普电梯有限公司 | Moving target recognition method and intelligent elevator billing system having moving target recognition function |
CN104318240A (en) * | 2014-07-18 | 2015-01-28 | 天津理工大学 | Flower bud discriminating method based on computer vision |
CN104506820A (en) * | 2015-01-06 | 2015-04-08 | 成都新舟锐视科技有限公司 | Intelligent multi-target tracing system |
CN104965964A (en) * | 2015-08-06 | 2015-10-07 | 山东建筑大学 | Building personnel distribution model establishment method based on surveillance video analysis |
CN105551266A (en) * | 2015-12-08 | 2016-05-04 | 合肥寰景信息技术有限公司 | Method of calculating pedestrian flow threshold of traffic signal controller |
CN105574499A (en) * | 2015-12-15 | 2016-05-11 | 东华大学 | Method and system for detecting and counting number of people based on SOC |
CN105912975A (en) * | 2015-02-20 | 2016-08-31 | 空中客车集团印度私人有限公司 | Management of aircraft in-cabin activities occuring during turnaround using video analytics |
CN106600590A (en) * | 2016-12-12 | 2017-04-26 | 四川华雁信息产业股份有限公司 | Automatic detection method and apparatus for oil leakage of current transformer |
CN106600576A (en) * | 2016-10-25 | 2017-04-26 | 上海萨戈电子科技有限公司 | RGB-D camera-based human head locking method |
CN106600643A (en) * | 2016-10-25 | 2017-04-26 | 长安大学 | People counting method based on trajectory analysis |
CN106846297A (en) * | 2016-12-21 | 2017-06-13 | 深圳市镭神智能系统有限公司 | Pedestrian's flow quantity detecting system and method based on laser radar |
CN106845620A (en) * | 2016-12-19 | 2017-06-13 | 江苏慧眼数据科技股份有限公司 | A kind of passenger flow counting method based on quene state analysis |
CN106845374A (en) * | 2017-01-06 | 2017-06-13 | 清华大学 | Pedestrian detection method and detection means based on deep learning |
CN107123126A (en) * | 2017-03-29 | 2017-09-01 | 天棣网络科技(上海)有限公司 | A kind of stream of people's moving scene temperature method of estimation |
CN107180420A (en) * | 2016-03-09 | 2017-09-19 | 顺丰科技有限公司 | Article transmits householder method |
CN107729804A (en) * | 2017-08-31 | 2018-02-23 | 广东数相智能科技有限公司 | A kind of people flow rate statistical method and device based on garment ornament |
CN107770921A (en) * | 2017-12-07 | 2018-03-06 | 四川谋席科技有限公司 | A kind of energy-saving LED illumination system based on Cloud Server |
CN107809834A (en) * | 2017-12-07 | 2018-03-16 | 四川谋席科技有限公司 | A kind of resident illuminator based on cloud computing |
WO2018059408A1 (en) * | 2016-09-29 | 2018-04-05 | 北京市商汤科技开发有限公司 | Cross-line counting method, and neural network training method and apparatus, and electronic device |
CN108229368A (en) * | 2017-12-28 | 2018-06-29 | 浙江大华技术股份有限公司 | A kind of image display method and device |
CN108549835A (en) * | 2018-03-08 | 2018-09-18 | 深圳市深网视界科技有限公司 | Crowd counts and its method, terminal device and the storage medium of model construction |
CN109389016A (en) * | 2017-08-10 | 2019-02-26 | 株式会社日立制作所 | A kind of method and system that the number of people counts |
CN109643321A (en) * | 2017-02-15 | 2019-04-16 | 富士通株式会社 | Video analytic system and video analysis method based on video monitoring |
CN109753842A (en) * | 2017-11-01 | 2019-05-14 | 深圳先进技术研究院 | A kind of method and device that flow of the people counts |
CN109800683A (en) * | 2018-12-30 | 2019-05-24 | 昆明物理研究所 | A kind of infrared pedestrian detection method and device based on FPGA |
CN110008766A (en) * | 2019-03-26 | 2019-07-12 | 南京大学 | A kind of occupancy statistical method based on Radio Frequency Identification Technology |
CN110826581A (en) * | 2018-08-10 | 2020-02-21 | 京东数字科技控股有限公司 | Animal number identification method, device, medium and electronic equipment |
CN111260628A (en) * | 2020-01-15 | 2020-06-09 | 北京林业大学 | Large nursery stock number counting method based on video image and electronic equipment |
CN111680569A (en) * | 2020-05-13 | 2020-09-18 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
CN112005247A (en) * | 2020-06-12 | 2020-11-27 | 深圳盈天下视觉科技有限公司 | People flow data monitoring system and people flow data display method and device thereof |
CN113536891A (en) * | 2021-05-10 | 2021-10-22 | 新疆爱华盈通信息技术有限公司 | Pedestrian traffic statistical method, storage medium and electronic equipment |
WO2022042352A1 (en) * | 2020-08-28 | 2022-03-03 | 安翰科技(武汉)股份有限公司 | Image recognition method, electronic device and readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6674877B1 (en) * | 2000-02-03 | 2004-01-06 | Microsoft Corporation | System and method for visually tracking occluded objects in real time |
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 |
CN101325690A (en) * | 2007-06-12 | 2008-12-17 | 上海正电科技发展有限公司 | Method and system for detecting human flow analysis and crowd accumulation process of monitoring video flow |
-
2011
- 2011-12-16 CN CN201110423349.2A patent/CN102542289B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6674877B1 (en) * | 2000-02-03 | 2004-01-06 | Microsoft Corporation | System and method for visually tracking occluded objects in real time |
CN101325690A (en) * | 2007-06-12 | 2008-12-17 | 上海正电科技发展有限公司 | Method and system for detecting human flow analysis and crowd accumulation process of monitoring video flow |
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 |
Non-Patent Citations (1)
Title |
---|
高陈强等: "视频图像中基于特征匹配的人流量统计", 《电视技术》 * |
Cited By (49)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103839035A (en) * | 2012-11-22 | 2014-06-04 | 富士通株式会社 | Person number statistical method and person number statistical system |
CN103186903A (en) * | 2013-04-22 | 2013-07-03 | 苏州科技学院 | Moving object detecting and tracking system based on PID (Proportion Integration Differentiation) |
CN104239905A (en) * | 2013-06-17 | 2014-12-24 | 上海盖普电梯有限公司 | Moving target recognition method and intelligent elevator billing system having moving target recognition function |
CN103400113B (en) * | 2013-07-10 | 2016-08-24 | 重庆大学 | Freeway tunnel pedestrian detection method based on image procossing |
CN103400113A (en) * | 2013-07-10 | 2013-11-20 | 重庆大学 | Method for detecting pedestrian on expressway or in tunnel based on image processing |
CN104318240A (en) * | 2014-07-18 | 2015-01-28 | 天津理工大学 | Flower bud discriminating method based on computer vision |
CN104123569A (en) * | 2014-07-21 | 2014-10-29 | 中国人民解放军63908部队 | Video person number information statistics method based on supervised learning |
CN104506820A (en) * | 2015-01-06 | 2015-04-08 | 成都新舟锐视科技有限公司 | Intelligent multi-target tracing system |
CN105912975B (en) * | 2015-02-20 | 2021-10-26 | 空中客车集团印度私人有限公司 | Managing aircraft in-cabin behavior occurring during a station crossing using video analysis |
CN105912975A (en) * | 2015-02-20 | 2016-08-31 | 空中客车集团印度私人有限公司 | Management of aircraft in-cabin activities occuring during turnaround using video analytics |
CN104965964A (en) * | 2015-08-06 | 2015-10-07 | 山东建筑大学 | Building personnel distribution model establishment method based on surveillance video analysis |
CN104965964B (en) * | 2015-08-06 | 2018-01-30 | 山东建筑大学 | A kind of building personnel's distributed model method for building up based on monitor video analysis |
CN105551266A (en) * | 2015-12-08 | 2016-05-04 | 合肥寰景信息技术有限公司 | Method of calculating pedestrian flow threshold of traffic signal controller |
CN105574499A (en) * | 2015-12-15 | 2016-05-11 | 东华大学 | Method and system for detecting and counting number of people based on SOC |
CN105574499B (en) * | 2015-12-15 | 2019-08-20 | 东华大学 | A kind of number detection statistics method and system based on SOC |
CN107180420A (en) * | 2016-03-09 | 2017-09-19 | 顺丰科技有限公司 | Article transmits householder method |
WO2018059408A1 (en) * | 2016-09-29 | 2018-04-05 | 北京市商汤科技开发有限公司 | Cross-line counting method, and neural network training method and apparatus, and electronic device |
CN106600576A (en) * | 2016-10-25 | 2017-04-26 | 上海萨戈电子科技有限公司 | RGB-D camera-based human head locking method |
CN106600643A (en) * | 2016-10-25 | 2017-04-26 | 长安大学 | People counting method based on trajectory analysis |
CN106600590A (en) * | 2016-12-12 | 2017-04-26 | 四川华雁信息产业股份有限公司 | Automatic detection method and apparatus for oil leakage of current transformer |
CN106600590B (en) * | 2016-12-12 | 2019-09-06 | 四川华雁信息产业股份有限公司 | Current transformer oil leak automatic testing method and device |
CN106845620B (en) * | 2016-12-19 | 2019-09-10 | 江苏慧眼数据科技股份有限公司 | A kind of passenger flow counting method based on quene state analysis |
CN106845620A (en) * | 2016-12-19 | 2017-06-13 | 江苏慧眼数据科技股份有限公司 | A kind of passenger flow counting method based on quene state analysis |
CN106846297A (en) * | 2016-12-21 | 2017-06-13 | 深圳市镭神智能系统有限公司 | Pedestrian's flow quantity detecting system and method based on laser radar |
CN106845374B (en) * | 2017-01-06 | 2020-03-27 | 清华大学 | Pedestrian detection method and detection device based on deep learning |
CN106845374A (en) * | 2017-01-06 | 2017-06-13 | 清华大学 | Pedestrian detection method and detection means based on deep learning |
CN109643321A (en) * | 2017-02-15 | 2019-04-16 | 富士通株式会社 | Video analytic system and video analysis method based on video monitoring |
CN107123126A (en) * | 2017-03-29 | 2017-09-01 | 天棣网络科技(上海)有限公司 | A kind of stream of people's moving scene temperature method of estimation |
CN109389016A (en) * | 2017-08-10 | 2019-02-26 | 株式会社日立制作所 | A kind of method and system that the number of people counts |
CN109389016B (en) * | 2017-08-10 | 2023-04-07 | 株式会社日立制作所 | Method and system for counting human heads |
CN107729804A (en) * | 2017-08-31 | 2018-02-23 | 广东数相智能科技有限公司 | A kind of people flow rate statistical method and device based on garment ornament |
CN109753842A (en) * | 2017-11-01 | 2019-05-14 | 深圳先进技术研究院 | A kind of method and device that flow of the people counts |
CN107770921A (en) * | 2017-12-07 | 2018-03-06 | 四川谋席科技有限公司 | A kind of energy-saving LED illumination system based on Cloud Server |
CN107809834A (en) * | 2017-12-07 | 2018-03-16 | 四川谋席科技有限公司 | A kind of resident illuminator based on cloud computing |
CN108229368A (en) * | 2017-12-28 | 2018-06-29 | 浙江大华技术股份有限公司 | A kind of image display method and device |
CN108229368B (en) * | 2017-12-28 | 2020-05-26 | 浙江大华技术股份有限公司 | Video display method and device |
CN108549835A (en) * | 2018-03-08 | 2018-09-18 | 深圳市深网视界科技有限公司 | Crowd counts and its method, terminal device and the storage medium of model construction |
CN110826581A (en) * | 2018-08-10 | 2020-02-21 | 京东数字科技控股有限公司 | Animal number identification method, device, medium and electronic equipment |
CN110826581B (en) * | 2018-08-10 | 2023-11-07 | 京东科技控股股份有限公司 | Animal number identification method, device, medium and electronic equipment |
CN109800683A (en) * | 2018-12-30 | 2019-05-24 | 昆明物理研究所 | A kind of infrared pedestrian detection method and device based on FPGA |
CN110008766B (en) * | 2019-03-26 | 2020-04-24 | 南京大学 | Indoor people counting method based on radio frequency identification technology |
CN110008766A (en) * | 2019-03-26 | 2019-07-12 | 南京大学 | A kind of occupancy statistical method based on Radio Frequency Identification Technology |
CN111260628A (en) * | 2020-01-15 | 2020-06-09 | 北京林业大学 | Large nursery stock number counting method based on video image and electronic equipment |
CN111680569A (en) * | 2020-05-13 | 2020-09-18 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
CN111680569B (en) * | 2020-05-13 | 2024-04-19 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
CN112005247A (en) * | 2020-06-12 | 2020-11-27 | 深圳盈天下视觉科技有限公司 | People flow data monitoring system and people flow data display method and device thereof |
WO2022042352A1 (en) * | 2020-08-28 | 2022-03-03 | 安翰科技(武汉)股份有限公司 | Image recognition method, electronic device and readable storage medium |
CN113536891A (en) * | 2021-05-10 | 2021-10-22 | 新疆爱华盈通信息技术有限公司 | Pedestrian traffic statistical method, storage medium and electronic equipment |
CN113536891B (en) * | 2021-05-10 | 2023-01-03 | 新疆爱华盈通信息技术有限公司 | Pedestrian traffic statistical method, storage medium and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN102542289B (en) | 2014-06-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102542289B (en) | Pedestrian volume statistical method based on plurality of Gaussian counting models | |
CN108986064B (en) | People flow statistical method, equipment and system | |
CN107527009B (en) | Remnant detection method based on YOLO target detection | |
Asha et al. | Vehicle counting for traffic management system using YOLO and correlation filter | |
CN106447680B (en) | The object detecting and tracking method that radar is merged with vision under dynamic background environment | |
CN103824070B (en) | A kind of rapid pedestrian detection method based on computer vision | |
CN105224912B (en) | Video pedestrian's detect and track method based on movable information and Track association | |
CN101389004B (en) | Moving target classification method based on on-line study | |
US8798327B2 (en) | Method and system for people flow statistics | |
CN104637058B (en) | A kind of volume of the flow of passengers identify statistical methods based on image information | |
CN108985170A (en) | Transmission line of electricity hanger recognition methods based on Three image difference and deep learning | |
CN105469105A (en) | Cigarette smoke detection method based on video monitoring | |
CN104978567B (en) | Vehicle checking method based on scene classification | |
CN105046206B (en) | Based on the pedestrian detection method and device for moving prior information in video | |
CN106919902B (en) | Vehicle identification and track tracking method based on CNN | |
CN102521565A (en) | Garment identification method and system for low-resolution video | |
CN103310194A (en) | Method for detecting head and shoulders of pedestrian in video based on overhead pixel gradient direction | |
CN105513053B (en) | One kind is used for background modeling method in video analysis | |
CN104992453A (en) | Target tracking method under complicated background based on extreme learning machine | |
CN104835147A (en) | Method for detecting crowded people flow in real time based on three-dimensional depth map data | |
CN110287907B (en) | Object detection method and device | |
CN109918971A (en) | Number detection method and device in monitor video | |
CN103955949A (en) | Moving target detection method based on Mean-shift algorithm | |
CN109684986B (en) | Vehicle analysis method and system based on vehicle detection and tracking | |
CN104063692A (en) | Method and system for pedestrian positioning detection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |