CN105069816B - A kind of method and system of inlet and outlet people flow rate statistical - Google Patents
A kind of method and system of inlet and outlet people flow rate statistical Download PDFInfo
- Publication number
- CN105069816B CN105069816B CN201510453202.6A CN201510453202A CN105069816B CN 105069816 B CN105069816 B CN 105069816B CN 201510453202 A CN201510453202 A CN 201510453202A CN 105069816 B CN105069816 B CN 105069816B
- Authority
- CN
- China
- Prior art keywords
- people
- image
- feature
- color
- flow rate
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000011161 development Methods 0.000 claims abstract description 25
- 230000033001 locomotion Effects 0.000 claims abstract description 23
- 238000012216 screening Methods 0.000 claims abstract description 17
- 238000000605 extraction Methods 0.000 claims abstract description 13
- 230000037308 hair color Effects 0.000 claims description 12
- 238000003708 edge detection Methods 0.000 claims description 10
- 239000000284 extract Substances 0.000 claims description 8
- 238000005286 illumination Methods 0.000 claims description 8
- 230000004927 fusion Effects 0.000 claims description 7
- 239000003086 colorant Substances 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000018109 developmental process Effects 0.000 description 21
- 238000006073 displacement reaction Methods 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 4
- 238000009416 shuttering Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 238000000205 computational method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000010410 dusting Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of method and system of inlet and outlet people flow rate statistical.The method of the inlet and outlet people flow rate statistical, go out moving region according to Motion feature extraction, there is the region of black and light color development feature with the colour model extraction in color space, motion feature and color development feature are merged, the coloured image of the candidate number of people is obtained, edge extracting is carried out to the coloured image of the candidate number of people, obtains the contour feature of the candidate number of people, fine screening is carried out to the coloured image of the candidate number of people according to number of people contour feature, identifies the number of people and count tracking.The beneficial effects of the invention are as follows：Reduce the background with number of people feature under complex scene and the omission factor of number of people detection is reduced to the flase drop of the number of people, improves the accuracy of inlet and outlet people flow rate statistical.
Description
（One）Technical field
The invention belongs to video monitoring, Computer Vision and analysis, technical field of machine vision, more particularly to it is a kind of into
Export the method and system of people flow rate statistical.
（Two）Background technology
With the continuous development of science and technology and information system management, to supermarket, market, station, tourist attractions, bus, sport
The flow of the people supervision of the different occasion such as race, public place of entertainment, festival activity is more information-based, intelligent.By to different fields
The demographics of conjunction can be reasonable distribution service and pipe with real-time estimation flow of the people, analysis stream of people distribution, the estimation degree of crowding
Reason resource, scientific dispatch, safety guarantee provide reliable basis.
Existing people flow rate statistical includes mainly：Artificial statistical, infrared scan counts and the people based on video analysis
Number statistics.Artificial statistical in a short time, flow of the people it is sparse in the case of it is reliable, but with the lengthening of time and
The increase of stream of people's metric density, statistical accuracy will be greatly lowered.To the crowd is dense, counting error is big for infrared scan counting, because
Cannot distinguish one or more people for infrared scan counting, cannot distinguish between people into outgoing direction.People flow rate statistical based on video
The programming count of high density flow of the people may be implemented, the existing people flow rate statistical based on video is mainly the following method：
First, the method based on human body segmentation, this method extracts moving target by motion analysis first, then passes through
Target Segmentation method, the image segmentation such as based on edge, the Target Segmentation based on coordinate mapping, the target point based on K-Means
It the methods of cuts and moving target is split, single human body target is obtained, to realize the statistics of flow of the people.The disadvantages of this method
It is in the big scene of stream of people's metric density, the trunk of human body is easy to mutually block, and the segmentation of target is highly difficult, to influence people
The statistics of flow.
Second is that the method based on the number of people or head and shoulder, this method replaces the identification of human body with the method for the identification number of people or head and shoulder,
The occlusion issue of trunk can be effectively avoided, and the number of people is more closely similar to rigid-object, target detection is made to identify problem
Become simpler.There are the people flow rate statistical method detected based on the number of people, method one at present：Hair, face are extracted with half-tone information
Product size removes the inhuman head region in part, and carrying out edge detection to obtained binary image extracts profile, is carried out to profile
Hough transform identifies the number of people；Method two：It is identified with the method for moving object detection and Hough transform or several annulus templates
The number of people；Method three：It is combined with the colour information of YCrCb color spaces and three frame difference Extracting of Moving Object, extraction is candidate
Then number of people region is scanned with several circular shutterings and calculates the identification number of people with tight ness rating.Method one is maximum the disadvantage is that meeting
By object identification similar in gray scale in static background and hair at hair, a degree of flase drop is caused, while half-tone information does not have
Chromatic colour abundant information will also result in certain influence.The shortcomings that method two be Hough transform can by color be different from color development but
It is to have the componental movement target identification of similar round feature at the number of people, such as the posture of the shoulder of people or certain moving target.
Several annulus templates not only can color be different from color development but the componental movement target identification with similar round feature is at the number of people,
And the equally distributed information of number of people profile can be lost, cause the outline identification for being segmented profile point aggregation in annulus at the number of people.Side
Three ratio method one of method and method two are all improved, but the template that method three uses is simpler, and several circular shutterings are only to searching
The region of rope is defined, and determines whether the number of people only with this standard of tight ness rating.Region is carried out with several circular shutterings
Scanning, and can first select to meet the large area of condition as the number of people, further accurate differentiation, can make the false drop rate of identification
It increases, while several circular shutterings have blindness during search is matched, if only carrying out the number of people with tight ness rating calculating
Identification can cause the imperfect of the identification number of people, to influence the accuracy of identification.
（Three）Invention content
In order to compensate for the shortcomings of the prior art, the present invention provides it is a kind of inlet and outlet people flow rate statistical method and system,
Number of people gray scale will easily be had just with the half-tone information and shape information of the number of people by solving current number of people detection algorithm in background
The problem of being judged as number of people region with the object of shape feature, reducing under complex scene has the background of number of people feature to the number of people
Flase drop, reduce the number of people detection omission factor, improve inlet and outlet people flow rate statistical accuracy.
The present invention is achieved through the following technical solutions：
A method of inlet and outlet people flow rate statistical, it is characterized in that：Include the following steps：
（1）, according to camera obtain image, go out moving target using Motion feature extraction；
（2）, using the colour model of color space, extract the region with black and light hair feature in image；
（3）, motion feature and color development feature merged, obtain the coloured image of the candidate number of people；
（4）, edge extracting is carried out to the coloured image of the candidate number of people, obtain the contour feature of the candidate number of people；
（5）, fine screening carried out to the coloured image of the candidate number of people according to the contour feature of the number of people, identify the number of people；
（6）, count tracking is carried out to the number of people that identifies.
Preferably, in step（1）In, by detecting region of variation in sequence image, moving target is carried from image
It takes out, while also needing in real time be updated background model.
Preferably, in step（2）In, according to black and light hair color regularity of distribution under different illumination, chooses and close
Suitable color space establishes the model of the distribution of color rule of description black hair and light hair in color space.
Preferably, in step（4）In, the coloured image of the candidate number of people is converted into gray level image, then to gray level image
Each pixel is detected, and is found out the larger point of gray-value variation, and handle these points, is made these points that can connect
Get up and constitute several lines, obtains the contour feature of the candidate number of people.
Preferably, in step（5）In, the number of people region of acquisition is the target of a similar round, by similar round target
Identification, realize and fine screening carried out to the coloured image of the candidate number of people, identify the number of people.
Preferably, in step（6）In, the number of people to identifying judges the disengaging of target by tracking result into line trace
Direction, and demographics are carried out in different directions.
A kind of system of inlet and outlet people flow rate statistical, it is characterized in that：Including Fusion Features modular system, the Fusion Features
Modular system is connected separately with moving target recognition modular system, color development characteristic extracting module system and edge detection module system
System, edge detection module system are connected with number of people screening module system, and number of people screening module system is connected with count tracking module
System.
The beneficial effects of the invention are as follows：First, by motion feature, moving target is extracted, is reduced under complex scene
The influence that static background with number of people feature detects the number of people.Secondly, black and shallow is carried out to image with color space
The extraction of color hair, color model colour gamut is broad, and the color that human eye can perceive can be transferred through color model table
Reveal and；Color model is compared with gray scale, the color characteristic of description hair that more can be careful, for black hair and light color
Hair can extract, and keep the recognition effect of the number of people more preferable.Again, to the coloured image with color development feature and motion feature
The identification for carrying out class circle object, the marginal information that bianry image extraction is compared to the marginal information of coloured image extraction are richer
Richness, the number of people detected can be more, to improve the accuracy rate of number of people detection.
（Four）Description of the drawings
The present invention will be further described below with reference to the drawings.
Attached drawing 1 is the flow diagram of the present invention；
Attached drawing 2 is the structural schematic diagram of the present invention；
Attached drawing 3 is the flow diagram of the specific implementation method of demographics in the embodiment of the present invention；
Attached drawing 4 is the particular flow sheet of edge detection in attached drawing 3；
Attached drawing 5 is the particular flow sheet of Head recognition in attached drawing 3.
（Five）Specific implementation mode
Attached drawing is a kind of specific embodiment of the present invention.The embodiment includes the following steps：（1）, obtained according to camera
Image goes out moving target using Motion feature extraction；（2）, using the colour model of color space, extracting has black in image
With the region of light hair feature；（3）, motion feature and color development feature merged, obtain the coloured image of the candidate number of people；
（4）, edge extracting is carried out to the coloured image of the candidate number of people, obtain the contour feature of the candidate number of people；（5）, according to the wheel of the number of people
Wide feature carries out fine screening to the coloured image of the candidate number of people, identifies the number of people；（6）, to the number of people that identifies into line trace meter
Number.In step（1）In, by detecting region of variation in sequence image, moving target is extracted from image, simultaneously also
It needs in real time to be updated background model.In step（2）In, according to black and light hair distribution of color under different illumination
Rule chooses suitable color space, and the distribution of color rule of description black hair and light hair is established in color space
Model.In step（4）In, the coloured image of the candidate number of people is converted into gray level image, then to each pixel of gray level image
Point is detected, and is found out the larger point of gray-value variation, and handle these points, is made these points that can connect composition
Several lines obtain the contour feature of the candidate number of people.In step（5）In, the number of people region of acquisition is the target of a similar round,
By the identification to similar round target, realizes and fine screening is carried out to the coloured image of the candidate number of people, identify the number of people.In step
（6）In, to the number of people that identifies into line trace, by tracking result judge target into outgoing direction, and in different directions into
Row demographics.
A kind of method of inlet and outlet people flow rate statistical using the present invention, is as follows：
Step 101, the image obtained according to camera, goes out moving target using Motion feature extraction.
Step 102, using the colour model of color space, the area with black and light hair feature in image is extracted
Domain.
Step 103, motion feature and color development feature are merged, obtains the coloured image of the candidate number of people.
Step 104, edge extracting is carried out to the coloured image of the candidate number of people, obtains the contour feature of the candidate number of people.
Step 105, fine screening is carried out to the coloured image of the candidate number of people according to the contour feature of the number of people, identifies the number of people.
Step 106, count tracking is carried out to the number of people identified.
Moving target is obtained using following method in step 101：
In the case of camera is fixed, background is between successive frame will not be changed, and the object only moved is
Image can be made to change, be modeled according to the color value of each pixel in image, by current image and background model into
Certain compares row, if the pixel color value in the pixel color value and background model on present image coordinate in corresponding coordinate has
When larger difference, current pixel is considered as foreground, is otherwise background, determines foreground target according to the result of the comparison.Meanwhile it pressing
Background is updated according to certain learning rate, to meet real-time demand.
In step 102, region of the extraction with black and light hair feature by taking Lab color spaces as an example：
Color characteristic has rotational invariance, direction insensitive, extracts relatively easy quick.Influence of the illumination to color can
To be made up by acquiring largely sample with black and light hair under different illumination, a large amount of different illumination conditions are acquired
Image with the number of people down, samples image color development pixel, and count it in the distribution situation of Lab color spaces, builds
Vertical hair color model.Each pixel color value in current image is compared with hair color model, if the picture of present image
Plain color value meets the condition of hair color model, then it is assumed that is color development, is otherwise non-hair color, obtains having according to the result of the comparison black
The region of color and light hair feature.
In step 103, the fusion of motion feature and color development feature is the fusion in Pixel-level.By two with motion feature
Value image and binary image with color development feature carry out operation, are had the binary image there are two types of feature simultaneously.
Computational methods are as follows：
Wherein, F (x, y) be pixel value of the binary image with motion feature at (x, y) point, H (x, y) be with
Pixel value of the binary image of color development feature at (x, y) point, T (x, y) are while having the binary image there are two types of feature
Pixel value at (x, y).
It is the image of binaryzation by what algorithm above obtained, if carrying out edge detection to binary image, obtains
Profile information it is fewer, lost the detailed information in region, in order to enrich profile information, needed to convert binary image
For coloured image.So, it is also necessary to mask operation is carried out to binary image, computational methods are as follows：
Wherein, T (x, y) is while pixel value of the tool there are two types of the binary image of feature at (x, y), R (x, y) are to work as
Pixel value of the preceding coloured image at (x, y).
In step 104, abundant profile information extracts people by taking the Canny algorithms of optimal edge detection as an example in order to obtain
Head contour feature.Its step are as follows：
Step 401, noise is eliminated, convolution algorithm is carried out using Gaussian filter；
Step 402, gradient magnitude and direction are calculated, respectively in horizontal and vertical derivation, on both horizontally and vertically
Derivative calculations gradient magnitude and direction；
Step 403, non-maxima suppression excludes non-edge pixels, only retains the hachure of some candidate edges；
Step 404, hysteresis threshold connects, and is compared to the amplitude of each location of pixels using high threshold and Low threshold.
If the amplitude of a certain location of pixels is more than high threshold, which is just left edge pixel；If a certain location of pixels
Amplitude is less than Low threshold, which is excluded；If the amplitude of a certain location of pixels is between two thresholds, which only exists
It is retained when being connected to a pixel for being higher than high threshold.
In step 105, video camera is shot from top to bottom, and rigid-object is more closely similar to for the number of people and nonrigid human body,
According to the contour feature of the number of people, number of people region is the target of a similar round.By the identification to similar round target, realize to waiting
Choose head coloured image carry out fine screening, identify the number of people.By taking Hough gradient algorithm as an example, it is as follows：
Step 501, to each non-zero pixels of edge image, its position is marked, and calculate partial gradient；
Step 502, each point to add up on the specified straight line of slope, obtains accumulator image；
Step 503, during pixel value is more than threshold value and is candidate more than the center pixel of its all neighbour in accumulator image
Heart point, and descending arranges；
Step 504, according to it, the descending at a distance from candidate centers arranges edge image non-zero pixels, what selection was most supported
One radius；
Step 505, candidate centers obtain edge image non-zero pixels and most adequately support, and with it is selected before
Central point has enough distances, then candidate centers are retained, and obtain the center of circle and radius.
In step 106, since the displacement of the number of people target identified between front and back two frame will not change a lot, institute
With, whether coincidence is had according to area to determine whether being same target, and judge into outgoing direction according to the direction of displacement, it is real
The demographics of existing different directions.
A kind of method of inlet and outlet people flow rate statistical using the present invention, specific implementation method are as follows：
Step 301, the image with the number of people under a large amount of different illumination conditions is acquired, image color development pixel is adopted
Sample；
Step 302, it is counted in the distribution situation of color space, establishes hair color model；
Step 303, input picture is obtained from camera；
Step 304, it is modeled according to the color value of each pixel in image；
Step 305, if pixel face in pixel color value and background model on present image coordinate in corresponding coordinate
When color value has larger difference, current pixel is considered as foreground, is otherwise background, determines foreground target according to the result of the comparison；
Step 306, background is updated according to certain learning rate, to meet real-time demand；
Step 307, each pixel color value in current image is compared with hair color model, if present image
Pixel color value meet the condition of hair color model, then it is assumed that be color development, be otherwise non-hair color, had according to the result of the comparison
There is the region of black and light hair feature；
Step 308, the binary image with motion feature and the binary image with color development feature are carried out and is transported
It calculates, is had the binary image there are two types of feature simultaneously, using the binary image as mask, mask fortune is carried out to present image
It calculates, obtains the coloured image of the candidate number of people；
Step 309, edge extracting is carried out to the coloured image of the candidate number of people, obtains the contour feature of the candidate number of people；
Step 310, according to the contour feature of the number of people, number of people region is the target of a similar round.By to similar round mesh
Target identifies, realizes and carries out fine screening to the coloured image of the candidate number of people, identifies the number of people；
Step 311, it will not be changed a lot, be utilized according to the displacement of the number of people target identified between front and back two frame
Whether area has coincidence to determine whether being same target, and judges, into outgoing direction, to realize different according to the direction of displacement
The demographics in direction.
A kind of system of inlet and outlet people flow rate statistical using the present invention, including：
Moving target recognition modular system, the image for being obtained according to camera go out movement using Motion feature extraction
Target；
Color development characteristic extracting module system, for using color space colour model, extract image in have black and
The region of light hair feature；
Fusion Features modular system obtains the colour of the candidate number of people for merging motion feature and color development feature
Image；
Edge detection module system carries out edge extracting for the coloured image to the candidate number of people, obtains the candidate number of people
Contour feature；
Number of people screening module system, for carrying out dusting cover to the coloured image of the candidate number of people according to the contour feature of the number of people
Choosing, identifies the number of people；
Count tracking modular system, for carrying out count tracking to the number of people identified.
In moving target recognition modular system 201, specifically for being modeled according to the color value of each pixel in image,
Compared with current image is carried out certain with background model, if in pixel color value and background model on present image coordinate
When pixel color value in corresponding coordinate has larger difference, current pixel is considered as foreground, is otherwise background, according to what is compared
As a result foreground target is determined.Meanwhile background is updated according to certain learning rate, to meet real-time demand.
In color development characteristic extracting module system 202, it is specifically used for acquiring the figure with the number of people under a large amount of different illumination conditions
Picture samples image color development pixel, and counts it in the distribution situation of color space, establishes hair color model.According to working as
The comparison result of each pixel color value and hair color model in preceding image obtains the area with black and light hair feature
Domain.
In Fusion Features modular system 203, it is specifically used for by the binary image with motion feature and with color development spy
The binary image of sign carries out and operation, is had the binary image there are two types of feature simultaneously, is to cover with the binary image
Code carries out mask operation to present image, obtains the coloured image of the candidate number of people.
In edge detection module system 204, specifically for the coloured image of the candidate number of people is converted to gray level image, then
The each pixel of gray level image is detected, the larger point of gray-value variation is found out, and handle these points, makes these
Point, which can connect, constitutes several lines, obtains the contour feature of the candidate number of people.
In number of people screening module system 205, it is specifically used for the contour feature according to the number of people, number of people region is a similar round
Target.By the identification to similar round target, realizes and fine screening is carried out to the coloured image of the candidate number of people, identify the number of people.
In count tracking modular system 206, it will not be occurred according to the displacement of the number of people target identified between front and back two frame
Prodigious variation, using area whether have coincidence to determine whether be same target, and according to the direction of displacement come judge into
Outgoing direction realizes the demographics of different directions.
These embodiments are not limitation of the present invention, according to function, method or knot made by these embodiments
Equivalent transformation on structure or replacement, all belong to the scope of protection of the present invention within.
A series of detailed description listed by the present invention, only specifically to the feasible embodiment of the present invention
It is bright, not to limit the scope of the invention, without departing substantially from spirit of that invention or essential characteristic, with others
Concrete form, equivalent way, change mode realize the present invention, should all be included in the protection scope of the present invention.
The present invention describes by way of example, but be not each embodiment only include an independent technical side
Case should also consider the specification as a whole, and the technical solutions in the various embodiments may also be suitably combined, form this field
The other embodiment that technical staff is appreciated that.
In addition, the embodiment of the present invention is described with flowchart and/or the block diagram, computer program instructions implementation process
Figure and/or block diagram, in addition to can providing method, system（Device）Or outside computer program product, computer program can also be provided
It instructs in computer Embedded Processor or other programmable data processing devices, it is made to generate flow chart and/or box
Function in figure.
Claims (4)
1. a kind of method of inlet and outlet people flow rate statistical, it is characterized in that：Include the following steps：
（1）, according to camera obtain image, go out moving target using Motion feature extraction；
（2）, using the colour model of color space, extract the region with black and light hair feature in image：According to black
Color and light hair color regularity of distribution under different illumination, get colors space, and description black head is established in color space
The model of the distribution of color rule of hair and light hair；
（3）, motion feature and color development feature merged, obtain the coloured image of the candidate number of people；
（4）, edge extracting is carried out to the coloured image of the candidate number of people, obtain the contour feature of the candidate number of people：By the candidate number of people
Coloured image is converted to gray level image, is then detected to each pixel of gray level image, it is larger to find out gray-value variation
Point, and these points are handled, make these points that can connect and constitute several lines, the profile for obtaining the candidate number of people is special
Sign；
（5）, fine screening carried out to the coloured image of the candidate number of people according to the contour feature of the number of people, identify the number of people：The people of acquisition
Head region is the target of a similar round, by the identification to similar round target, realizes and is carried out to the coloured image of the candidate number of people
Fine screening identifies the number of people；
（6）, count tracking is carried out to the number of people that identifies.
2. a kind of method of inlet and outlet people flow rate statistical according to claim 1, it is characterized in that：In step（1）In, pass through
Region of variation is detected in sequence image, and moving target is extracted from image, while also needing in real time to background model
It is updated.
3. a kind of method of inlet and outlet people flow rate statistical according to claim 1, it is characterized in that：In step（6）In, to knowing
The number of people not gone out into line trace, by tracking result judge target into outgoing direction, and carry out demographics in different directions.
4. the system used in a kind of method of inlet and outlet people flow rate statistical as described in claim 1, it is characterized in that：Melt including feature
Block system is molded, the Fusion Features modular system is connected separately with moving target recognition modular system, color development feature extraction mould
Block system and edge detection module system, edge detection module system are connected with number of people screening module system, number of people screening module
System is connected with count tracking modular system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510453202.6A CN105069816B (en) | 2015-07-29 | 2015-07-29 | A kind of method and system of inlet and outlet people flow rate statistical |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510453202.6A CN105069816B (en) | 2015-07-29 | 2015-07-29 | A kind of method and system of inlet and outlet people flow rate statistical |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105069816A CN105069816A (en) | 2015-11-18 |
| CN105069816B true CN105069816B (en) | 2018-09-11 |
Family
ID=54499172
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510453202.6A Active CN105069816B (en) | 2015-07-29 | 2015-07-29 | A kind of method and system of inlet and outlet people flow rate statistical |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105069816B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845318B (en) * | 2015-12-03 | 2019-06-21 | 杭州海康威视数字技术股份有限公司 | Passenger flow information acquisition method and device, passenger flow information processing method and processing device |
| CN106028160A (en) * | 2016-06-03 | 2016-10-12 | 腾讯科技（深圳）有限公司 | Image data processing method and device |
| CN107766784A (en) * | 2016-08-20 | 2018-03-06 | 宋坤骏 | A kind of novel video people counting algorithm |
| CN108109117A (en) * | 2017-12-14 | 2018-06-01 | 柳州智视科技有限公司 | A kind of method of the color real-time transform based on moving object |
| CN109101866B (en) * | 2018-06-05 | 2020-12-15 | 中国科学院自动化研究所 | Pedestrian re-identification method and system based on segmentation silhouette |
| CN114355468A (en) * | 2021-12-31 | 2022-04-15 | 深圳市汇顶科技股份有限公司 | Test method and test device for fingerprint detection device |
| CN117132948B (en) * | 2023-10-27 | 2024-01-30 | 南昌理工学院 | Scenic area tourist flow monitoring method, system, readable storage medium and computer |
-
2015
- 2015-07-29 CN CN201510453202.6A patent/CN105069816B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 基于计算机视觉的人流量双向统计;王瑞 等;《电子技术应用》;20120930;第38卷(第9期);摘要、第1、2节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105069816A (en) | 2015-11-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105069816B (en) | A kind of method and system of inlet and outlet people flow rate statistical | |
| CN115082419B (en) | Blow-molded luggage production defect detection method | |
| CN108520226B (en) | Pedestrian re-identification method based on body decomposition and significance detection | |
| CN104166841B (en) | The quick detection recognition methods of pedestrian or vehicle is specified in a kind of video surveillance network | |
| CN108288033B (en) | A kind of safety cap detection method based on random fern fusion multiple features | |
| CN105160319B (en) | One kind realizing that pedestrian knows method for distinguishing again under monitor video | |
| WO2022099598A1 (en) | Video dynamic target detection method based on relative statistical features of image pixels | |
| CN108629319B (en) | Image detection method and system | |
| CN108537239B (en) | Method for detecting image saliency target | |
| CN103605953B (en) | Vehicle interest target detection method based on sliding window search | |
| CN109918971B (en) | Method and device for detecting number of people in monitoring video | |
| CN105046197B (en) | Multi-template pedestrian detection method based on cluster | |
| CN102867188B (en) | Method for detecting seat state in meeting place based on cascade structure | |
| CN104077605B (en) | A kind of pedestrian's search recognition methods based on color topological structure | |
| CN104978567B (en) | Vehicle checking method based on scene classification | |
| CN106874884B (en) | Human body re-identification method based on part segmentation | |
| CN105631892B (en) | It is a kind of that detection method is damaged based on the aviation image building of shade and textural characteristics | |
| CN105893925A (en) | Human hand detection method based on complexion and device | |
| CN107273832B (en) | License plate recognition method and system based on integral channel characteristics and convolutional neural network | |
| CN104268583A (en) | Pedestrian re-recognition method and system based on color area features | |
| CN104715238A (en) | Pedestrian detection method based on multi-feature fusion | |
| CN102194108A (en) | Smiley face expression recognition method based on clustering linear discriminant analysis of feature selection | |
| CN102214309A (en) | Special human body recognition method based on head and shoulder model | |
| CN102915433A (en) | Character combination-based license plate positioning and identifying method | |
| CN109271932A (en) | Pedestrian based on color-match recognition methods again |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200113 Address after: 1202, block a, building 3, phase I, Tian'an Yungu Industrial Park, Bantian street, Longgang District, Shenzhen City, Guangdong Province Patentee after: Shenzhen jovision Technology Co. Ltd. Address before: Xinluo Avenue high tech Zone of Ji'nan City, Shandong province 250000 silver bearing No. 2008 Building No. 4-1101 Patentee before: Jovision Technology Co., Ltd. |
|
| TR01 | Transfer of patent right |