CN109034124A - A kind of intelligent control method and system - Google Patents
A kind of intelligent control method and system Download PDFInfo
- Publication number
- CN109034124A CN109034124A CN201811003619.2A CN201811003619A CN109034124A CN 109034124 A CN109034124 A CN 109034124A CN 201811003619 A CN201811003619 A CN 201811003619A CN 109034124 A CN109034124 A CN 109034124A
- Authority
- CN
- China
- Prior art keywords
- image
- human body
- pedestrian
- monitoring image
- crowd massing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Alarm Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of intelligent control method and system, the intelligent control methods, comprising: S1, obtains monitoring image, and delimit warning region in the monitoring image；S2, pedestrian detection is carried out to the monitoring image, and exports pedestrian detection image；S3, human body attitude estimation is carried out to the pedestrian detection image, and export the key message image including human body key message；S4, using foot's key point of the human body in the key message image as position of human body point, and when judging that the position of human body point is in the warning region, record the monitoring image at the moment.The present invention is monitored pedestrian using pedestrian detection and human body attitude estimation by delimiting warning region to the monitoring image of acquisition, judge monitoring image warning region whether someone, carry out initiative alarming to realize and invade the pedestrian of important area.
Description
Technical field
The invention belongs to field of intelligent monitoring, in particular to a kind of intelligent control method and system.
Background technique
Intelligent monitoring is to be integrated with intelligent behavior recognizer in embedded video server, can be in pictured scene
Pedestrian or the behavior of vehicle identify, judge, and under suitable condition, generate warning note user.Although existing at present
School, office building, residential quarters, the environmental applications such as prison intelligent monitor system, but existing intelligent monitor system cannot
Enter for important area someone and carry out initiative alarming, it is desired nonetheless to which the Security Personnel for guarding monitoring carries out monitoring in 24 hours
It can just note abnormalities, it would therefore be highly desirable to propose a kind of intelligent monitoring that can carry out pedestrian detection and initiative alarming to important area
System,
Summary of the invention
It is an object of the invention to: that cannot enter aiming at the problem that carry out initiative alarming important area someone, mention
For a kind of intelligent control method and system, by delimiting warning region, while using pedestrian detection and human body attitude estimation to row
People is monitored, and carries out initiative alarming to realize and invade the pedestrian of important area.
The technical solution adopted by the invention is as follows:
A kind of intelligent control method, comprising:
S1, monitoring image is obtained, and delimit warning region in the monitoring image；
S2, pedestrian detection is carried out to the monitoring image, and export the pedestrian detection image including pedestrian's frame；
S3, human body attitude estimation is carried out to pedestrian's frame in the pedestrian detection image, and exporting includes that human body key is believed
The key message image of breath；
S4, using foot's key point of the human body in the key message image as position of human body point, and described in the judgement
When position of human body point is in the warning region, the monitoring image is recorded.
Further, step S2 specifically:
S21, YOLOv2 network frame, training pedestrian detection model are based on；
S22, pedestrian detection is carried out to the monitoring image using trained pedestrian detection model, and exporting includes pedestrian
The pedestrian detection image of frame.
Further, the training set of the pedestrian detection model is Caltech Pedestrian Dataset pedestrian's data
Collection.
Further, step S3 specifically:
S31, double branch multistage convolutional neural networks are established；
S32, pedestrian's frame in the pedestrian detection image is returned using double branch multistage convolutional neural networks
Return, obtain include connection between human body key point and human body key point thermal map；
S33, on the link division to each pedestrian between the human body key point and human body key point in the thermal map,
Then refine is carried out, output includes the connection key message image between human body key point and human body key point.
Further, the intelligent control method, further includes:
S5, it is carried out based on the step S1 monitoring image obtained or the key message image obtained based on step S3
Crowd massing judgement, and when judgement has crowd massing, record the monitoring image at the moment.
Further, in step S5, crowd massing judgement is carried out based on the obtained key message image of step S3
Process, specifically:
(1), using foot's key point of the human body in the key message image as position of human body point, the monitoring is counted
Foot's key point number in the warning region of image obtains warning region and monitors number；
(2), when warning region monitoring number is more than the maximum number threshold limit of setting, it is judged as the security area
There is crowd massing in domain.
Further, in step S5, the process of crowd massing judgement is carried out based on the monitoring image that step S1 is obtained,
Specifically:
(1), the monitoring image is handled, generates density feature figure；
(2), following formula is used to the density feature figure, counts crowd massing degree:
Wherein, L, W are respectively the length and width of the density feature figure, wherein zl,wRefer to that density generated is special
Levy the pixel value size in figure in the position (l, w) pixel, CiIt refers to as the crowd massing degree in the monitoring image；
(3), when the crowd massing degree is more than the maximum crowd massing degree threshold value of setting, it is judged as in the warning region
There is crowd massing.
A kind of intelligent monitor system, comprising:
Image Acquisition and processing unit delimit warning region for obtaining monitoring image, and in the monitoring image；
Pedestrian detection unit for carrying out pedestrian detection to the monitoring image, and exports the pedestrian including pedestrian's frame and examines
Altimetric image；
Human body attitude estimation unit, for carrying out human body attitude estimation to pedestrian's frame in the pedestrian detection image, and
Output includes the key message image of human body key message；
Region intrusion alarm unit, for using foot's key point of the human body in the key message image as human body position
It sets a little, and when judging that the position of human body point is in the warning region, records the monitoring image.
Further, the intelligent monitor system, further includes:
Crowd massing alarm unit for the monitoring image based on described image acquisition and processing unit acquisition or is based on institute
It states the key message image that human body attitude estimation unit obtains and carries out crowd massing judgement.
Further, the crowd massing alarm unit, comprising: the first clustering collection alarm unit or the second crowd massing
Alarm unit；
The first clustering collection alarm unit, the crucial letter for being obtained based on the human body attitude estimation unit
It ceases image and carries out crowd massing judgement；
The second crowd massing alarm unit, the monitoring image for being obtained based on described image acquisition with processing unit
Carry out crowd massing judgement.
In conclusion by adopting the above-described technical solution, the beneficial effects of the present invention are:
1, the present invention be by delimiting warning region to the monitoring image of acquisition, while being estimated using pedestrian detection and human body attitude
Meter is monitored pedestrian, judge monitoring image warning region whether someone, the pedestrian of important area is invaded to realize
Carry out initiative alarming.
2, the present invention is based on YOLOv2 network frames, are made using Caltech Pedestrian Dataset pedestrian's data set
For training set, the pedestrian detection model trained can predict the image of a variety of scales with mandatory learning, to improve pedestrian detection
Model is to the robustness of the pedestrian detection of the picture of different scale size, and so as to improve the performance of algorithmic system.
3, the present invention passes through only to human body attitude estimation is carried out in the region of pedestrian's frame in pedestrian detection image, to drop
The low calculation amount of algorithm, improves the efficiency of system, ensure that the real-time of system.
4, crowd density estimation of the invention is suitable for a variety of actual environments.It may determine that in monitoring image or security area
Crowd massing situation in domain, the place excessive to crowd massing carry out effective early warning.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is the flow chart of intelligent control method of the invention.
Fig. 2 is the effect diagram of the warning region of monitoring image of the invention.
Fig. 3 is the structure chart of YOLOv2 network frame of the invention.
Fig. 4 is the structure chart of the YOLOv2 network frame of the invention by taking the image of 416 × 416 sizes of input as an example.
Fig. 5 is the structure chart of double branch multistage convolutional neural networks of the invention.
Fig. 6 is the effect of key message image of the present invention including the connection between human body key point and human body key point
Schematic diagram.
Fig. 7 is density feature illustrated example of the invention.
Fig. 8 is the structural block diagram of intelligent monitor system of the invention.
Marked in the figure: 10- Image Acquisition and processing unit, 20- pedestrian detection unit, 30- human body attitude estimation unit,
The region 40- intrusion alarm unit, the first clustering collection alarm unit of 51-, 52- the second crowd massing alarm unit.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention, i.e., described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is logical
The component for the embodiment of the present invention being often described and illustrated herein in the accompanying drawings can be arranged and be designed with a variety of different configurations.
Therefore, the detailed description of the embodiment of the present invention provided in the accompanying drawings is not intended to limit below claimed
The scope of the present invention, but be merely representative of selected embodiment of the invention.Based on the embodiment of the present invention, those skilled in the art
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
It should be noted that the relational terms of term " first " and " second " or the like be used merely to an entity or
Operation is distinguished with another entity or operation, and without necessarily requiring or implying between these entities or operation, there are any
This actual relationship or sequence.Moreover, the terms "include", "comprise" or its any other variant be intended to it is non-exclusive
Property include so that include a series of elements process, method, article or equipment not only include those elements, but also
Further include other elements that are not explicitly listed, or further include for this process, method, article or equipment it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described
There is also other identical elements in the process, method, article or equipment of element.
A kind of intelligent control method, comprising:
S1, monitoring image is obtained, and delimit warning region in the monitoring image；
S2, pedestrian detection is carried out to the monitoring image, and export the pedestrian detection image including pedestrian's frame；
S3, human body attitude estimation is carried out to pedestrian's frame in the pedestrian detection image, and exporting includes that human body key is believed
The key message image of breath；
S4, using foot's key point of the human body in the key message image as position of human body point, and described in the judgement
When position of human body point is in the warning region, the monitoring image is recorded.
A kind of intelligent monitor system, comprising:
Image Acquisition and processing unit 10, delimit warning region for obtaining monitoring image, and in the monitoring image；
Pedestrian detection unit 20 for carrying out pedestrian detection to the monitoring image, and exports the pedestrian including pedestrian's frame
Detection image；
Human body attitude estimation unit 30, for carrying out human body attitude estimation to pedestrian's frame in the pedestrian detection image,
And export the key message image including human body key message；
Region intrusion alarm unit 40, for using foot's key point of the human body in the key message image as human body
Location point, and when judging that the position of human body point is in the warning region, record the monitoring image.
The present invention is estimated by delimiting warning region to the monitoring image of acquisition using pedestrian detection and human body attitude
Pedestrian is monitored, judge monitoring image warning region whether someone, thus realize to the pedestrian of important area invade into
Row initiative alarming.
Feature and performance of the invention are described in further detail with reference to embodiments.
Embodiment 1
A kind of intelligent control method, as shown in Figure 1, comprising:
S1, monitoring image is obtained, and delimit warning region in the monitoring image；
By taking a certain monitoring system as an example, monitoring image is obtained by monitoring camera, when practical application, monitoring camera is obtained
Taking is real-time video picture, the important area in the video pictures, and being delimited by one identification frame of mark is warning region,
Such as inlet and outlet, main thoroughfare or corridor mouth of cell of company etc., the effect diagram of warning region as shown in Fig. 2,
It will be arranged at the channel on many places doorway in warning region.The notation methods of the warning region can be and delimit the police by software
Guarding against region indicates that each frame monitoring image of the video pictures has this stroke by delimiting warning region in monitoring image
Fixed warning region.And it is possible to be delimited manually by Security Personnel, it can also change at any time, by the demand specifically monitored
Depending on.
S2, pedestrian detection is carried out to the monitoring image, and export the pedestrian detection image including pedestrian's frame；
In terms of security protection, the robustness of pedestrian detection largely affects the other function module of security system
Operate normally the overall performance with security system.Therefore, the pedestrian detection that the present embodiment uses, specifically includes:
S21, YOLOv2 network frame, training pedestrian detection model are based on；
The YOLOv2 network frame is the network of a full convolution, and structure chart is as shown in Figure 3.The pedestrian detection mould
The training set of type is Caltech Pedestrian Dataset pedestrian's data set.
The size of the input picture of the YOLOv2 network frame determines the size of output image, big with input 416 × 416
For small image, as shown in figure 4, the image size of its output is 13 × 13, in hands-on, training set is handled as ruler
Very little size is { 320 × 320,352 × 352 ... ..., 608 × 608 }, step-length 32, after 10 iteration of every progress, from training
It concentrates the image for randomly choosing a kind of size to be again trained as input, the figure of a variety of scales can be predicted with mandatory learning
Picture, thus improve pedestrian detection model to the robustness of the pedestrian detection of the picture of different scale size, and so as to improve algorithm
The performance of system.
S22, pedestrian detection is carried out to the monitoring image using trained pedestrian detection model, and exporting includes pedestrian
The pedestrian detection image of frame.Identify the pedestrian in monitoring image, and by the way that pedestrian's frame is arranged to each pedestrian, it will be each
Pedestrian marks out from monitoring image to come.
S3, human body attitude estimation is carried out to pedestrian's frame in the pedestrian detection image, and exporting includes that human body key is believed
The key message image of breath；
It specifically includes:
S31, double branch multistage convolutional neural networks are established；
The structure of double branch multistage convolutional neural networks is as shown in figure 5, include one Branch1 of branch and branch two
Branch2；
S32, the pedestrian detection image is returned using double branch multistage convolutional neural networks, including
The thermal map of connection between human body key point and human body key point；
Specifically, the two-dimentional confidence atlas S that partes corporis humani divides location information is exported by one Branch1 of branch, there are J to set
Letter figure, a corresponding confidence map per anthropoid key point, to carry out the detection of human body key point；
S=S1,S2,…,SJ)SJ∈RW×H, J ∈ { 1,2 ..., J }
And body parts affinity vector field L is exported by two Branch2 of branch, there are C vector field, every class limbs pair
A vector field is answered, to carry out the detection of the connection of human body key point.
L=(L1,L2,…,LC)LC∈RW×H×2, C ∈ { 1,2 ..., C }
S33, on the link division to each pedestrian between the human body key point and human body key point in the thermal map,
Then refine is carried out, output includes the key message image of the connection between human body key point and human body key point, effect
Schematic diagram is as shown in Figure 6.
S4, using foot's key point of the human body in the key message image as position of human body point, and described in the judgement
When position of human body point is in the warning region, the monitoring image is recorded；
In the key message image including the connection between human body key point and human body key point, the head of pedestrian,
The major joints such as four limbs point is noted as human body key point, can therefrom find foot's key point as position of human body point,
Detect that foot's key point of human body judges someone in the warning region in warning region, to alarm.
Further, the intelligent control method, further includes:
S5, it is carried out based on the step S1 monitoring image obtained or the key message image obtained based on step S3
Crowd massing judgement, and when judgement has crowd massing, record the monitoring image at the moment.It should be understood that crowd massing
The monitoring image that the monitoring image and warning region intrusion alarm of alarm logging are recorded need to carry out different marks to distinguish report
Alert type.When needed, also exportable warning message, such as voice signal prompt Security Personnel.
For small range visual angle scene, can be counted by the number in the warning region to monitoring image, thus
The case where judging crowd massing carries out the process of crowd massing judgement based on the obtained key message image of step S3,
Specifically:
(1), using foot's key point of the human body in the key message image as position of human body point, the monitoring is counted
Foot's key point number in the warning region of image obtains warning region and monitors number；
(2), when warning region monitoring number is more than the maximum number threshold limit of setting, it is judged as the security area
There is crowd massing in domain.
And for large-scale visual angle scene, such as subway, square, street etc. needs to judge crowd to entire monitoring image
The case where aggregation, carries out the process of crowd massing judgement based on the monitoring image that step S1 is obtained, specifically:
(1), the monitoring image is handled, generates density feature figure；
Image can be handled using existing trained single channel convolutional neural networks, to generate such as Fig. 7
Shown in density feature figure.
(2), following formula is used to the density feature figure, counts crowd massing degree:
Wherein, L, W are respectively the length and width of the density feature figure, wherein zl,wRefer to that density generated is special
Levy the pixel value size in figure in the position (l, w) pixel, CiIt refers to as the crowd massing degree in the monitoring image.
(3), when the crowd massing degree is more than the maximum crowd massing degree threshold value of setting, it is judged as in the warning region
There is crowd massing.
Embodiment 2
Intelligent control method described in 1 in conjunction with the embodiments, the present embodiment provides a kind of intelligent monitor systems, comprising:
Image Acquisition and processing unit 10, delimit warning region for obtaining monitoring image, and in the monitoring image；
Pedestrian detection unit 20 for carrying out pedestrian detection to the monitoring image, and exports the pedestrian including pedestrian's frame
Detection image；
Human body attitude estimation unit 30, for carrying out human body attitude estimation to pedestrian's frame in the pedestrian detection image,
And export the key message image including human body key message；
Region intrusion alarm unit 40, for using foot's key point of the human body in the key message image as human body
Location point, and when judging that the position of human body point is in the warning region, record the monitoring image.
The intelligent monitor system, further includes:
Crowd massing alarm unit, for acquiring the monitoring image obtained with processing unit 10 based on described image or being based on
The key message image that the human body attitude estimation unit 30 obtains carries out crowd massing judgement.
The crowd massing alarm unit, comprising: the first clustering collection alarm unit 51 or the alarm of the second crowd massing are single
Member 52；
The first clustering collection alarm unit 51, the pass for being obtained based on the human body attitude estimation unit 30
Key information image carries out crowd massing judgement；
The second crowd massing alarm unit 52, the monitoring for being obtained based on described image acquisition with processing unit 10
Image carries out crowd massing judgement.
It is apparent to those skilled in the art that the convenience and letter for description are bought, the intelligence of foregoing description
The specific work process of monitoring system and its each functional unit, can be with reference in the intelligent control method in previous embodiment 1
Corresponding process, details are not described herein.
Above-mentioned bright each functional unit can integrate in one processing unit, is also possible to the independent physics of each unit and deposits
It can also be integrated in one unit with two or more units.Above-mentioned integrated unit can both use the shape of hardware
Formula is realized, can also be realized in the form of software functional units.
If integrated each functional unit is realized in the form of SFU software functional unit and sells as independent product
Or it in use, can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention
Substantially all or part of the part that contributes to existing technology or the technical solution can be with software product in other words
Form embody, which is stored in a storage medium, including some instructions use so that one
Computer equipment (can be smart phone, tablet computer, personal computer, server or the network equipment etc.) executes this hair
The all or part of the steps of the bright intelligent control method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only storage
Device (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or light
The various media that can store program code such as disk.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (10)
1. a kind of intelligent control method characterized by comprising
S1, monitoring image is obtained, and delimit warning region in the monitoring image；
S2, pedestrian detection is carried out to the monitoring image, and export the pedestrian detection image including pedestrian's frame；
S3, human body attitude estimation is carried out to pedestrian's frame in the pedestrian detection image, and exporting includes human body key message
Key message image；
S4, using foot's key point of the human body in the key message image as position of human body point, and judging the human body
When location point is in the warning region, the monitoring image at the moment is recorded.
2. intelligent control method as described in claim 1, which is characterized in that step S2 specifically:
S21, YOLOv2 network frame, training pedestrian detection model are based on；
S22, pedestrian detection is carried out to the monitoring image using trained pedestrian detection model, and exporting includes pedestrian's frame
Pedestrian detection image.
3. intelligent control method as claimed in claim 2, which is characterized in that the training set of the pedestrian detection model is
Caltech Pedestrian Dataset pedestrian's data set.
4. intelligent control method as described in claim 1, which is characterized in that step S3 specifically:
S31, double branch multistage convolutional neural networks are established；
S32, pedestrian's frame in the pedestrian detection image is returned using double branch multistage convolutional neural networks, is obtained
To the thermal map including the connection between human body key point and human body key point；
S33, on the link division to each pedestrian between the human body key point and human body key point in the thermal map, then
Refine is carried out, output includes the connection key message image between human body key point and human body key point.
5. intelligent control method as described in claim 1, which is characterized in that further include:
S5, crowd is carried out based on the step S1 monitoring image obtained or the key message image obtained based on step S3
Aggregation judgement, and when judgement has crowd massing, record the monitoring image at the moment.
6. intelligent control method as claimed in claim 5, which is characterized in that in step S5, obtained based on step S3 described in
Key message image carries out the process of crowd massing judgement, specifically:
(1), using foot's key point of the human body in the key message image as position of human body point, the monitoring image is counted
Warning region in foot's key point number, obtain warning region monitor number；
(2), when judging that the warning region monitoring number is more than the maximum number threshold limit of setting, it is judged as the warning region
Inside there is crowd massing.
7. intelligent control method as claimed in claim 5, which is characterized in that in step S5, based on described in step S1 acquisition
Monitoring image carries out the process of crowd massing judgement, specifically:
(1), the monitoring image is handled, generates density feature figure；
(2), following formula is used to the density feature figure, counts crowd massing degree:
Wherein, L, W are respectively the length and width of the density feature figure, wherein zl,wRefer to density feature figure generated
In the position (l, w) pixel pixel value size, CiIt refers to as the crowd massing degree in the monitoring image；
(3), judge that the crowd massing degree is more than the maximum crowd massing degree threshold value of setting, be judged as someone in the warning region
Clustering collection.
8. a kind of intelligent monitor system characterized by comprising
Image Acquisition and processing unit (10), delimit warning region for obtaining monitoring image, and in the monitoring image；
Pedestrian detection unit (20) for carrying out pedestrian detection to the monitoring image, and exports the pedestrian including pedestrian's frame and examines
Altimetric image；
Human body attitude estimation unit (30), for carrying out human body attitude estimation to pedestrian's frame in the pedestrian detection image, and
Output includes the key message image of human body key message；
Region intrusion alarm unit (40), for using foot's key point of the human body in the key message image as human body position
It sets a little, and when judging that the position of human body point is in the warning region, records the monitoring image.
9. intelligent monitor system as claimed in claim 8, which is characterized in that further include:
Crowd massing alarm unit for the monitoring image based on described image acquisition and processing unit (10) acquisition or is based on institute
It states the key message image that human body attitude estimation unit (30) obtains and carries out crowd massing judgement.
10. intelligent monitor system as claimed in claim 9, which is characterized in that the crowd massing alarm unit, comprising: the
One crowd massing alarm unit (51) or the second crowd massing alarm unit (52)；
The first clustering collection alarm unit (51), the pass for being obtained based on the human body attitude estimation unit (30)
Key information image carries out crowd massing judgement；
The second crowd massing alarm unit (52), the monitoring for being obtained based on described image acquisition with processing unit (10)
Image carries out crowd massing judgement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811003619.2A CN109034124A (en) | 2018-08-30 | 2018-08-30 | A kind of intelligent control method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811003619.2A CN109034124A (en) | 2018-08-30 | 2018-08-30 | A kind of intelligent control method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109034124A true CN109034124A (en) | 2018-12-18 |
Family
ID=64625820
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811003619.2A Pending CN109034124A (en) | 2018-08-30 | 2018-08-30 | A kind of intelligent control method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109034124A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685041A (en) * | 2019-01-23 | 2019-04-26 | 北京市商汤科技开发有限公司 | Image analysis method and device, electronic equipment and storage medium |
| CN109753891A (en) * | 2018-12-19 | 2019-05-14 | 山东师范大学 | Football player's orientation calibration method and system based on human body critical point detection |
| CN110110657A (en) * | 2019-05-07 | 2019-08-09 | 中冶赛迪重庆信息技术有限公司 | Method for early warning, device, equipment and the storage medium of visual identity danger |
| CN110443984A (en) * | 2019-06-27 | 2019-11-12 | 维沃移动通信有限公司 | A kind of monitoring method and mobile terminal |
| CN110796032A (en) * | 2019-10-11 | 2020-02-14 | 深圳市誉托科技有限公司 | Video fence based on human body posture assessment and early warning method |
| CN110991375A (en) * | 2019-12-10 | 2020-04-10 | 北京航空航天大学 | Group behavior analysis method and device |
| CN111126317A (en) * | 2019-12-26 | 2020-05-08 | 腾讯科技（深圳）有限公司 | Image processing method, device, server and storage medium |
| CN111126257A (en) * | 2019-12-23 | 2020-05-08 | 上海商汤智能科技有限公司 | Behavior detection method and device |
| CN111209814A (en) * | 2019-12-27 | 2020-05-29 | 广东德融汇科技有限公司 | Face recognition campus area early warning method and system for K12 education stage |
| CN111246167A (en) * | 2020-01-13 | 2020-06-05 | 惠龙易通国际物流股份有限公司 | Electronic monitoring area setting method and device |
| CN111382726A (en) * | 2020-04-01 | 2020-07-07 | 浙江大华技术股份有限公司 | Engineering operation detection method and related device |
| CN111476277A (en) * | 2020-03-20 | 2020-07-31 | 广东光速智能设备有限公司 | Alarm method and system based on image recognition |
| CN111723598A (en) * | 2019-03-18 | 2020-09-29 | 北京邦天信息技术有限公司 | Machine vision system and implementation method thereof |
| CN111814587A (en) * | 2020-06-18 | 2020-10-23 | 浙江大华技术股份有限公司 | Human behavior detection method, teacher behavior detection method, and related system and device |
| CN113554837A (en) * | 2021-07-21 | 2021-10-26 | 山东浪潮通软信息科技有限公司 | Monitoring alarm method and system |
| CN114359875A (en) * | 2022-03-17 | 2022-04-15 | 成都考拉悠然科技有限公司 | Vehicle effective snapshot method and system |
| CN114662521A (en) * | 2021-11-16 | 2022-06-24 | 成都考拉悠然科技有限公司 | Method and system for detecting wandering behavior of pedestrian |
| CN116189311A (en) * | 2023-04-27 | 2023-05-30 | 成都愚创科技有限公司 | Protective clothing wears standardized flow monitoring system |
| CN117876973A (en) * | 2024-03-13 | 2024-04-12 | 中科盛阳信息技术有限公司 | Deep learning-based park security monitoring method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101835035A (en) * | 2010-06-04 | 2010-09-15 | 天津市亚安科技电子有限公司 | Regional invasion real-time detection method |
| KR20130067842A (en) * | 2011-12-14 | 2013-06-25 | 한국전자통신연구원 | Image based object information detecting method and apparatus |
| CN107911653A (en) * | 2017-11-16 | 2018-04-13 | 王磊 | The module of intelligent video monitoring in institute, system, method and storage medium |
| CN107944327A (en) * | 2016-10-10 | 2018-04-20 | 杭州海康威视数字技术股份有限公司 | A kind of demographic method and device |
| CN108009473A (en) * | 2017-10-31 | 2018-05-08 | 深圳大学 | Based on goal behavior attribute video structural processing method, system and storage device |
| CN108009466A (en) * | 2016-10-28 | 2018-05-08 | 北京旷视科技有限公司 | Pedestrian detection method and device |
| CN108205870A (en) * | 2016-12-16 | 2018-06-26 | 北京迪科达科技有限公司 | A kind of campus personal safety intelligent monitoring management system |
-
2018
- 2018-08-30 CN CN201811003619.2A patent/CN109034124A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101835035A (en) * | 2010-06-04 | 2010-09-15 | 天津市亚安科技电子有限公司 | Regional invasion real-time detection method |
| KR20130067842A (en) * | 2011-12-14 | 2013-06-25 | 한국전자통신연구원 | Image based object information detecting method and apparatus |
| CN107944327A (en) * | 2016-10-10 | 2018-04-20 | 杭州海康威视数字技术股份有限公司 | A kind of demographic method and device |
| CN108009466A (en) * | 2016-10-28 | 2018-05-08 | 北京旷视科技有限公司 | Pedestrian detection method and device |
| CN108205870A (en) * | 2016-12-16 | 2018-06-26 | 北京迪科达科技有限公司 | A kind of campus personal safety intelligent monitoring management system |
| CN108009473A (en) * | 2017-10-31 | 2018-05-08 | 深圳大学 | Based on goal behavior attribute video structural processing method, system and storage device |
| CN107911653A (en) * | 2017-11-16 | 2018-04-13 | 王磊 | The module of intelligent video monitoring in institute, system, method and storage medium |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109753891A (en) * | 2018-12-19 | 2019-05-14 | 山东师范大学 | Football player's orientation calibration method and system based on human body critical point detection |
| CN109685041A (en) * | 2019-01-23 | 2019-04-26 | 北京市商汤科技开发有限公司 | Image analysis method and device, electronic equipment and storage medium |
| CN111723598A (en) * | 2019-03-18 | 2020-09-29 | 北京邦天信息技术有限公司 | Machine vision system and implementation method thereof |
| CN110110657A (en) * | 2019-05-07 | 2019-08-09 | 中冶赛迪重庆信息技术有限公司 | Method for early warning, device, equipment and the storage medium of visual identity danger |
| CN110443984A (en) * | 2019-06-27 | 2019-11-12 | 维沃移动通信有限公司 | A kind of monitoring method and mobile terminal |
| CN110796032A (en) * | 2019-10-11 | 2020-02-14 | 深圳市誉托科技有限公司 | Video fence based on human body posture assessment and early warning method |
| CN110991375A (en) * | 2019-12-10 | 2020-04-10 | 北京航空航天大学 | Group behavior analysis method and device |
| CN111126257A (en) * | 2019-12-23 | 2020-05-08 | 上海商汤智能科技有限公司 | Behavior detection method and device |
| CN111126257B (en) * | 2019-12-23 | 2023-08-11 | 上海商汤智能科技有限公司 | Behavior detection method and device |
| CN111126317A (en) * | 2019-12-26 | 2020-05-08 | 腾讯科技（深圳）有限公司 | Image processing method, device, server and storage medium |
| CN111209814A (en) * | 2019-12-27 | 2020-05-29 | 广东德融汇科技有限公司 | Face recognition campus area early warning method and system for K12 education stage |
| CN111209814B (en) * | 2019-12-27 | 2023-05-19 | 广东德融汇科技有限公司 | Face recognition campus area early warning method and system for K12 education stage |
| CN111246167A (en) * | 2020-01-13 | 2020-06-05 | 惠龙易通国际物流股份有限公司 | Electronic monitoring area setting method and device |
| CN111476277A (en) * | 2020-03-20 | 2020-07-31 | 广东光速智能设备有限公司 | Alarm method and system based on image recognition |
| CN111382726B (en) * | 2020-04-01 | 2023-09-01 | 浙江大华技术股份有限公司 | Engineering operation detection method and related device |
| CN111382726A (en) * | 2020-04-01 | 2020-07-07 | 浙江大华技术股份有限公司 | Engineering operation detection method and related device |
| CN111814587A (en) * | 2020-06-18 | 2020-10-23 | 浙江大华技术股份有限公司 | Human behavior detection method, teacher behavior detection method, and related system and device |
| CN113554837A (en) * | 2021-07-21 | 2021-10-26 | 山东浪潮通软信息科技有限公司 | Monitoring alarm method and system |
| CN113554837B (en) * | 2021-07-21 | 2023-08-15 | 浪潮通用软件有限公司 | Monitoring alarm method and system |
| CN114662521A (en) * | 2021-11-16 | 2022-06-24 | 成都考拉悠然科技有限公司 | Method and system for detecting wandering behavior of pedestrian |
| CN114359875B (en) * | 2022-03-17 | 2022-05-17 | 成都考拉悠然科技有限公司 | Vehicle effective snapshot method and system |
| CN114359875A (en) * | 2022-03-17 | 2022-04-15 | 成都考拉悠然科技有限公司 | Vehicle effective snapshot method and system |
| CN116189311A (en) * | 2023-04-27 | 2023-05-30 | 成都愚创科技有限公司 | Protective clothing wears standardized flow monitoring system |
| CN117876973A (en) * | 2024-03-13 | 2024-04-12 | 中科盛阳信息技术有限公司 | Deep learning-based park security monitoring method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034124A (en) | A kind of intelligent control method and system | |
| Zhang et al. | Video anomaly detection based on locality sensitive hashing filters | |
| US11295139B2 (en) | Human presence detection in edge devices | |
| US20180278894A1 (en) | Surveillance system | |
| CN103514694A (en) | Intrusion detection monitoring system | |
| US11175185B2 (en) | Method, apparatus and system for passive infrared sensor framework | |
| CN105405150A (en) | Abnormal behavior detection method and abnormal behavior detection device based fused characteristics | |
| CN109117771B (en) | System and method for detecting violence events in image based on anchor nodes | |
| CN103106394A (en) | Human body action recognition method in video surveillance | |
| CN109033979B (en) | Indoor pedestrian detection method based on WIFI and camera sensor decision-level fusion | |
| WO2022048572A1 (en) | Target identification method and apparatus, and electronic device | |
| CN113963301A (en) | Space-time feature fused video fire and smoke detection method and system | |
| Kalaivani et al. | Towards comprehensive understanding of event detection and video summarization approaches | |
| KR20160093253A (en) | Video based abnormal flow detection method and system | |
| Feris et al. | Case study: IBM smart surveillance system | |
| Yu et al. | Long-term identity-aware multi-person tracking for surveillance video summarization | |
| Gayathri et al. | Intelligent smart home security system: A deep learning approach | |
| Dogra et al. | Scene representation and anomalous activity detection using weighted region association graph | |
| CN108038451A (en) | Anomaly detection method and device | |
| Kalaivani et al. | Video event representation for abnormal event detection | |
| Zhou et al. | Rapid and robust traffic accident detection based on orientation map | |
| Anithadevi et al. | An Improved SSD Object Detection Algorithm For Safe Social Distancing and Face Mask Detection In Public Areas Through Intelligent Video Analytics | |
| Hubner et al. | Audio-video sensor fusion for the detection of security critical events in public spaces | |
| Li et al. | Region-based trajectory analysis for abnormal behaviour detection: a trial study for suicide detection and prevention | |
| WO2021199315A1 (en) | Monitoring device, monitoring method, and recording medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181218 |