CN106897716A - A kind of dormitory safety monitoring system and method - Google Patents
A kind of dormitory safety monitoring system and method Download PDFInfo
- Publication number
- CN106897716A CN106897716A CN201710287122.7A CN201710287122A CN106897716A CN 106897716 A CN106897716 A CN 106897716A CN 201710287122 A CN201710287122 A CN 201710287122A CN 106897716 A CN106897716 A CN 106897716A
- Authority
- CN
- China
- Prior art keywords
- dormitory
- identified
- safety monitoring
- current environment
- luminous intensity
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19639—Details of the system layout
- G08B13/19641—Multiple cameras having overlapping views on a single scene
- G08B13/19643—Multiple cameras having overlapping views on a single scene wherein the cameras play different roles, e.g. different resolution, different camera type, master-slave camera
-
- 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)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The invention discloses a kind of dormitory safety monitoring system and method, current environment luminous intensity is monitored, when current environment luminous intensity is higher than preset first threshold value, using the video image in white light camera collection dormitory monitor area；When current environment luminous intensity is not higher than preset first threshold value, using the video image in infrared camera collection dormitory monitor area；Video image to collecting is pre-processed；The facial image feature of object to be identified in video image after pretreatment is extracted, and facial image feature is matched with the standard form for prestoring, to judge whether object to be identified is dormitory member.Dormitory safety monitoring system provided by the present invention and method, IMAQ is carried out using infrared camera, be adapted to the situation of illumination wretched insufficiency, it is to avoid night or other illumination not enough scene human face identification accuracy it is relatively low, monitoring dynamics is smaller, influence dormitory moves in the problem of personal security.
Description
Technical field
The present invention relates to safety monitoring technology field, more particularly to a kind of dormitory safety monitoring system and method.
Background technology
In today that dormitory is increasingly taken seriously safely, the uncertain factor that night non-dormitory personnel discrepancy dormitory brings
Very big potential safety hazard will be brought.Therefore, carrying out effective monitoring to dormitory at dead of night is beneficial to protect the life and wealth of student
Produce safety.But during the late into the night dormitory management personnel due to factors such as physiology, environment, often to the strange personnel's susceptibility in periphery not
Enough, the personnel to the dormitory that comes in and goes out are badly in need of being identified using more intelligentized method, to monitor the periphery suspicion of dormitory doorway
Doubt people.
Face recognition technology is the face feature based on human body, to the facial image or video flowing that are input into, judges it
With the presence or absence of face, the positional information of the position, size and each major facial organ of face is further identified, and according to this
The identity characteristic that is contained in a little each face of information extraction, and these template identity characteristics are contrasted, so as to recognize its body
Part information.
Existing face recognition technology is aimed at the more uniform sufficient situation of daylight mostly, and due to illumination not
Large effect can be uniformly produced to the identification of face checking, therefore, the knowledge of the even unglazed situation human face of dim light at night
Other accuracy is relatively low, the dynamics that can thus influence dormitory to monitor, and threatens dormitory to move in the life and property safety of personnel.
The content of the invention
It is an object of the invention to provide a kind of dormitory safety monitoring system and method, to solve existing dormitory monitoring technology
In the not enough scene human face identification accuracy of night or other illumination it is relatively low, monitoring dynamics is smaller, influence dormitory is moved in
The problem of personal security.
In order to solve the above technical problems, the present invention provides a kind of dormitory safety monitoring system, including：
White light camera, infrared camera, rotatable mechanical head framework, central processing unit and image processor；
Wherein, the white light camera and the infrared camera may be contained within the mechanical head framework, described
White light camera is used for when current environment luminous intensity is higher than preset first threshold value, the video figure in collection dormitory monitor area
Picture；The infrared camera is used to, when current environment luminous intensity is not higher than the preset first threshold value, gather dormitory monitored space
Video image in domain；
The central processing unit is connected with the white light camera and the infrared camera, for the institute to collecting
State video image to be pre-processed, and be input into described image processor；
Described image processor is connected with the central processing unit, wait to know in video image after pretreatment for extracting
The facial image feature of other object, and the facial image feature is matched with the standard form for prestoring, to judge
Whether the object to be identified is dormitory member.
Alternatively, also include：
The infrared light-feeding instrument being connected with the infrared camera, for current residing for the infrared camera when detecting
When ambient light intensity is less than default Second Threshold, light filling operation is carried out.
Alternatively, the white light camera and the infrared camera are autozoom camera, for when described
When image processor judges the object to be identified as non-dormitory member, object to be identified described in auto-focusing, and treated to described
The behavior of identification object is tracked monitoring.
Alternatively, also include：
The warning device being connected with described image processor, for judging the object to be identified when described image processor
During for non-dormitory member, warning reminding is carried out in the form of voice and/or electric bell.
Present invention also offers a kind of dormitory method for safety monitoring, including：
Monitoring current environment luminous intensity, when the current environment luminous intensity is higher than preset first threshold value, is taken the photograph using white light
As the video image in head collection dormitory monitor area；When the current environment luminous intensity is not higher than the preset first threshold value
When, using the video image in infrared camera collection dormitory monitor area；
The video image to collecting is pre-processed；
The facial image feature of object to be identified in video image after pretreatment is extracted, and the facial image is special
Levy and matched with the standard form for prestoring, to judge whether the object to be identified is dormitory member.
Alternatively, it is described that the facial image feature is matched with the standard form for prestoring, it is described to judge
Whether object to be identified is that dormitory member includes：
When the ambient light intensity is not higher than the preset first threshold value, dormitory is recognized using dynamic fluid recognition methods
Moving target in monitor area；
After facial image is recognized, facial zone is divided into many sub-regions；
Characteristic point and Corner Detection are carried out to each facial subregion using image processor, double attribute mathematical models are calculated
Similarity；
When the similarity after matching exceedes default similarity threshold, judge that the object to be identified is dormitory member；It is no
Then judge that the object to be identified is non-dormitory member.
Alternatively, it is described facial zone is divided into many sub-regions before also include：
Facial image to recognizing carries out horizontal motion correction.
Alternatively, also include：
When the object to be identified is judged as non-dormitory member, object to be identified described in auto-focusing is waited to know to described
The behavior of other object is tracked monitoring.
Alternatively, also include：
When the object to be identified is judged as non-dormitory member, warning reminding is carried out in the form of voice and/or electric bell.
Alternatively, also include：
When the current environment luminous intensity is detected less than default Second Threshold, light filling operation is carried out.
Dormitory safety monitoring system provided by the present invention and method, monitor current environment luminous intensity, in current environment light
When intensity is higher than preset first threshold value, using the video image in white light camera collection dormitory monitor area；In current environment
When luminous intensity is not higher than preset first threshold value, using the video image in infrared camera collection dormitory monitor area；To collection
To video image pre-processed；The facial image feature of object to be identified in video image after pretreatment is extracted, and
Facial image feature is matched with the standard form for prestoring, to judge whether object to be identified is dormitory member.This
The there is provided dormitory safety monitoring system of invention and method, fully apply in dormitory safety now emerging face recognition technology
Early warning system in, under the dark situations such as dormitory night, IMAQ is carried out using the camera of near-infrared, be adapted to illumination it is tight
The situation that weight is not enough, it is to avoid night or other illumination not enough scene human face identification accuracy is relatively low, monitoring dynamics
Smaller, influence dormitory moves in the problem of personal security.
The application can also be further provided with the infrared light-feeding instrument being connected with infrared camera, when detecting infrared photography
When current environment luminous intensity residing for head is less than default Second Threshold, light filling operation is carried out.By such setting, can be in ring
The border luminous intensity low-down late into the night ensures the abundance of illumination, to be monitored to dormitory surrounding enviroment.
White light camera and infrared camera are autozoom camera in the application, for sentencing when image processor
When determining object to be identified for non-dormitory member, object to be identified described in auto-focusing, and the behavior of object to be identified is carried out with
Track is monitored.By using white light camera and infrared camera with autozoom, being capable of auto-focusing monitoring juridical-person
The behavior of member.
Brief description of the drawings
Fig. 1 is a kind of structured flowchart of specific embodiment of dormitory safety monitoring system provided by the present invention；
Fig. 2 is a kind of specific embodiment flow chart of dormitory method for safety monitoring provided in an embodiment of the present invention；
Fig. 3 is another specific embodiment flow chart of dormitory method for safety monitoring provided in an embodiment of the present invention.Fig. 1
In：White light camera -1, infrared camera -2, rotatable mechanical head framework -3, central processing unit -4, image procossing
Device -5
Specific embodiment
In order that those skilled in the art more fully understand the present invention program, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.Obviously, described embodiment is only a part of embodiment of the invention, rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creative work premise
Lower obtained every other embodiment, belongs to the scope of protection of the invention.
A kind of structured flowchart of the specific embodiment of dormitory safety monitoring system provided by the present invention as shown in figure 1,
The system includes：
White light camera 1, infrared camera 2, rotatable mechanical head framework 3, central processing unit 4 and image procossing
Device 5；
Wherein, the white light camera 1 and the infrared camera 2 may be contained within the mechanical head framework 3, institute
White light camera 1 is stated for when current environment luminous intensity is higher than preset first threshold value, gathering the video in dormitory monitor area
Image；The infrared camera 2 is used for when current environment luminous intensity is not higher than the preset first threshold value, collection dormitory monitoring
Video image in region；
The central processing unit 4 is used to pre-process the video image for collecting, and is input into described image
Processor 5；
The facial image that described image processor 5 is used to extract object to be identified in video image after pretreatment is special
Levy, and the facial image feature is matched with the standard form for prestoring, whether to judge the object to be identified
It is dormitory member.
Mechanical head framework can drive white light camera and infrared camera to rotate in the application, be carried out with to target
Real-time tracing.
Further, dormitory safety monitoring system provided by the present invention can further include：Infrared light-feeding instrument, uses
In when the current environment luminous intensity residing for the infrared camera 2 is detected less than default Second Threshold, light filling operation is carried out.
It is pointed out that white light camera 1 and infrared camera 2 can be taken the photograph for autozoom in the embodiment of the present invention
It is to be identified described in auto-focusing for when described image processor judges the object to be identified as non-dormitory member as head
Object, and behavior to the object to be identified is tracked monitoring.
On the basis of any of the above-described embodiment, the present invention can further include：Also include：
Warning device, for when described image processor 4 judges the object to be identified as non-dormitory member, with voice
And/or the form of electric bell carries out warning reminding.
Additionally, image processor can specifically include multiple parallel computation units in the application, each parallel computation unit
Collaboration is processed, and can accelerate the speed for the treatment of.
The application can further include：Display screen, for being shown in real time to the video image for collecting.
Shown in real time by the current video image for collecting, dormitory supervisor can be easy to look into current environment
See, further improve the security of dormitory environment, it is ensured that the dynamics of monitoring.
If the similarity after system matching has exceeded threshold value, personage to be identified is the personnel in dormitory, then open
Gate inhibition allows it to enter, and then shows it is non-dormitory personnel after it fails to match, and white light camera 1 or infrared camera 2 adjust camera lens, from
The behavior of dynamic focusing monitoring unauthorized person, and administrative staff are reminded in the form of electric bell and/or voice on backstage.By such
Set, further increase the security of monitoring.Dormitory safety monitoring system provided by the present invention, monitors current environment light intensity
Degree, when current environment luminous intensity is higher than preset first threshold value, using the video in white light camera collection dormitory monitor area
Image；When current environment luminous intensity is not higher than preset first threshold value, using in infrared camera collection dormitory monitor area
Video image；Video image to collecting is pre-processed；Extract object to be identified in video image after pretreatment
Facial image feature, and facial image feature is matched with the standard form for prestoring, to judge that object to be identified is
No is dormitory member.Dormitory safety monitoring system provided by the present invention, fully by now emerging face recognition technology application
In the early warning system of dormitory safety, under the dark situations such as dormitory night, IMAQ is carried out using infrared camera, be adapted to
The situation of illumination wretched insufficiency, it is to avoid night or other illumination not enough scene human face identification accuracy is relatively low, monitoring
Dynamics is smaller, influence dormitory moves in the problem of personal security.
Dormitory method for safety monitoring provided in an embodiment of the present invention is introduced below, dormitory described below is supervised safely
Prosecutor method can be mutually to should refer to above-described dormitory safety monitoring system.
Fig. 2 is the flow chart of dormitory method for safety monitoring provided in an embodiment of the present invention, the dormitory security monitoring side of reference picture 2
Method can be specifically included：
Step S101：Monitoring current environment luminous intensity, when the current environment luminous intensity is higher than preset first threshold value, adopts
The video image in dormitory monitor area is gathered with white light camera；When the current environment luminous intensity is not higher than described default the
During one threshold value, using the video image in infrared camera collection dormitory monitor area；
Step S102：The video image to collecting is pre-processed；
Step S103：The facial image feature of object to be identified in video image after pretreatment is extracted, and will be described
Facial image feature is matched with the standard form for prestoring, to judge whether the object to be identified is dormitory member.
Dormitory method for safety monitoring provided by the present invention, monitors current environment luminous intensity, high in current environment luminous intensity
When preset first threshold value, using the video image in white light camera collection dormitory monitor area；In current environment luminous intensity
Not higher than preset first threshold value when, using infrared camera collection dormitory monitor area in video image；To regarding for collecting
Frequency image is pre-processed；Extract the facial image feature of object to be identified in video image after pretreatment, and by face
Characteristics of image is matched with the standard form for prestoring, to judge whether object to be identified is dormitory member.Institute of the present invention
The dormitory method for safety monitoring of offer, fully applies the early warning system in dormitory safety by now emerging face recognition technology
In, under the dark situations such as dormitory night, IMAQ is carried out using infrared camera, it is adapted to the situation of illumination wretched insufficiency,
Avoid and recognize that accuracy is relatively low, monitoring dynamics is smaller, influence dormitory at night or the not enough scene human face of other illumination
Move in the problem of personal security.
Embedded human face identification in the prior art runs the algorithm of recognition of face by CPU mostly, due to embedded system
The hardware resource of system is mostly limited, so recognition of face is less efficient.The embodiment of the present invention accelerates current people using GPU
Face recognizer, can greatly speed up the speed of service of face identification system.Simultaneously common lens are used in reality mostly, it is impossible to
Combine with face recognition technology and realize autozoom, so, this hair very limited to the behavior monitoring ability of suspect
Bright embodiment is capable of auto-focusing monitoring unauthorized person using white light camera and infrared camera with autozoom
Behavior.Additionally, fully the monitoring range of camera can be mentioned most using mechanical rotatable head framework in the present embodiment
Greatly, and according to the regional location of recognition of face realize autozoom, facilitating backstage to the monitoring of non-dormitory suspicion personnel and
Early warning.
Reference picture 3, the specific implementation with reference to concrete scene to dormitory safety monitoring system provided by the present invention
Cheng Jinhang is further elaborated on, and the process includes：
Step S200：The advance typing dormitory information about firms of face identification system, and store in the equipment of embedded system；
Step S201：Mechanical cradle head control camera level and upper and lower two and multi-angle in face recognition device system
Lower monitoring dormitory neighboring area；
Step S202：Open white light camera 1 daytime, night opens infrared camera 2, and the late into the night opens infrared light-feeding instrument,
Obtain the monitor video of dormitory porch；
Step S203：The pretreatment of common white light and near-infrared human face is carried out in embedded system family chip CPU；
Step S204：The feature of above-mentioned pretreated facial image is extracted in the CPU module of embedded system；
Step S205：Split facial zone by characteristic point areas, it is double using each piece of facial zone of GPU parallel computations
The similarity of the matching of attributed graph；
Step S206：If the similarity after matching has exceeded threshold value in system, then target to be identified is in dormitory
Personnel, opening gate；
Step S207：Then show target to be identified after it fails to match for non-dormitory personnel, camera adjustment camera lens is automatic right
The behavior of Jiao's monitoring unauthorized person, and administrative staff are reminded in the form of voice and/or electric bell on backstage.
The preprocessing process that CPU carries out recognition of face is illustrated below.
First it is the pretreatment to face in the case of daytime, the process includes：Skin cluster based on human body,
On the basis of construction good person's face complexion model, optimum under current environment is extracted using the machine learning algorithm of SVMs
Face on the colour of skin HSI numerical value.
Normally split hair, eyes and the lip region of human body again on the basis of the region for extracting face.In herein
In order that the eyes that must be separated, lip can accurately be recognized, it is necessary to extract eyes and lip in the x direction
Corroded, expanded in y-direction, made its signature, it is easy to recognize.
Finally carry out the judgement of the pre-identification of face.
Template matching method can specifically be used：The image of standard form and region to be matched is overlaped, master die
The pixel number of the rectangular area of plate is s, and overlapping region pixel number is designated as a, supplementary set of the overlapping region in whole rectangular area
Pixel number be b, matching degree is γ, then：If the value of γ is bigger, then show that matching degree is higher, otherwise
The value of γ is smaller, then show that matching degree is lower.
Elementary contour with face as template, using the matching algorithm of above-mentioned template, to the face binary map for extracting
As being matched, the region of low matching is eliminated by threshold determination method, screen the approximate region of face.
Centered on face, eyes, the lip proportionate relationship of similar triangles that is constituted each other judges whether to be
Meet the matching requirement of face.
Substantial amounts of non-face interference region is eliminated by the pre-identification of above-mentioned face, discrimination reach 80% with
On, while judging efficiency is high, be suitable for background it is relatively simple under the embedded situation for waiting limited system resources.
Because dormitory is very rare in time-division at night face and more quiet under dormitory surrounding enviroment, illumination bar
Part environment is poor, now detects the illuminance of surrounding environment less than in the case of preset first threshold value, and system automatically turns on red
Outer camera 2, when ambient light intensity is detected less than default Second Threshold, presets Second Threshold and is less than preset first threshold value,
Open infrared light-feeding instrument to carry out automatic supplement ambient light illumination to assist in identifying and monitor, what is read from infrared camera 2 is red
Outer image, black white image is obtained after being processed by optical filter.It is the characteristics of for the evening stream of people and less mobiles, movable
Unauthorized person is very protruded in such a case, and the identification of dynamic fluid is transported in being particularly suitable for detecting environment in current background
Dynamic human region.
The current motion under the monitoring of infrared camera 2 of frame differential method treatment identification is carried out in the thread of embedded type CPU
Personage.
Background matrix：By γ1, γ2Current frame image is weighted as weight factor and background image two field picture is made before
It is sample mean matrix, by regulating and controlling weight factor γ1, γ2Size be can with regulating system moving object is recognized it is sensitive
Degree.
Then each frame image matrix subtracts background frame image matrix, then obtains domain transformation by Threshold segmentation
Pixel distribution, in rectangle fitting domain transformation, the heart draws a length of a+a of rectangle wherein0, a width of b+b0Field as follow-up knowledge
Other alternative area, wherein a, the length of side of b fitted rectangles.
The alternative area of matching degree face higher is obtained after by above-mentioned face pre-identification, under this algorithm
The demand of majority of case can be met, but for complex environment, have the generation of the interference region of approximate face, together
When, because face all can produce large effect under various cunning angular deflection states to face authentication effect, so
A large amount of various states, and different faces are trained for a long time by Adaboost algorithm in this module, and
And training result is kept, non-face region is rejected by adaboost pattern-recognitions, and use eyes on this basis
Identification, distance is a between positioning eyes, and the left and right width of face center of face is b, is eliminated for a ＞ b/2 on the basis of modeling
Fall the situation of deflection angle excessive side face, obtain normal facial image.
Because the inclination of face can affect to follow-up authentication, so needing to carry out that level is affine rectifys
Just, it is above-mentioned recognize eyes the step of after, facial image to level is rotated according to eyes axis and horizontal tilt angle θ.
Illumination in the case of daytime is sufficient in dormitory region, uses the classical face recognition algorithms that PCA converts can be with
Meet and require, but when at candlelight the time point and the time-division at night, because uneven illumination and low-light (level) can be produced to recognition of face
Very big influence, now needs to be compensated in recognizer and looks after the uneven influence for producing, and is become using classical Groba small echos
The robustness that can increase recognition of face to illumination is changed, good recognition effect is reached.Therefore using double attributed graph recognizers
The recognition of face in the case of two kinds can very well be met.Due to the limitation of embedded type CPU resource, so using emerging GPU simultaneously
Row processes to accelerate authentication speed.The following is GPU concurrent operations are recognized with reference to human face's characteristic point region segmentation
Principle.
If directly can produce the discrimination of face not high to contrast using the facial image in whole facial image and storehouse,
Recognition efficiency is also than relatively low, so now using the multi-method identification in characteristic point region, identification process is as follows：Obtaining face
After face image, 120 pieces of regions are roughly divided into human face region.Examined using the single computing unit of GPU in every piece of region
Survey characteristic point and angle point.68 point feature points in human face region scope are substantially obtained, comprising eyes, nose, lip etc. and face
The characteristic points such as profile.The ability of the parallel computation of GPU is made full use of simultaneously, is that each computing unit distribution of GPU is special at each
Levy and carried out in field a little PCA conversion and Groba wavelet transformations and calculate its characteristic vector similarity, and by related power
Weigh coefficient to draw the similarity in characteristic point region, finally use substantially 68 characteristic area similarities in face face scope
The method of linear superposition draws the whole similarity of face.
This method takes full advantage of the discrimination high and wavelet transformation of PCA to the insensitivity of illumination, makes discrimination
It is more accurate, while influence of the illumination to recognition of face is decreased, and it is single that each calculating is dispatched using GPU parallel computations
Unit has greatly accelerated the speed of face verification matching to Similarity Measure in characteristic point region, is especially suitable for dormitory at dusk at night
Between the time-division recognition of face.
It is of the invention fully to apply in the early warning system of dormitory safety now emerging face recognition technology, at dormitory night
Under the dark situations such as evening, using the infrared camera infrared light filling of collocation, it is adapted to the situation of illumination wretched insufficiency, while at algorithm
In reason, employ fluid identification and recognize in advance, reduce the burden of global search face, it is contemplated that the situation of uneven illumination, use
GPU concurrent operation combination human face characteristic point Segmentation parallel computations its similarities, substantially increases recognition speed, and to light
According to there is good robustness.In face identification system, mechanical rotatable head framework 3 can fully by camera
Monitoring range mentions maximum, and realizes autozoom according to the regional location of recognition of face, to facilitate backstage to dislike non-dormitory
The monitoring and early warning of the personnel of doubting.
Each embodiment is described by the way of progressive in this specification, and what each embodiment was stressed is and other
The difference of embodiment, between each embodiment same or similar part mutually referring to.For being filled disclosed in embodiment
For putting, because it is corresponded to the method disclosed in Example, so description is fairly simple, related part is referring to method part
Illustrate.
Professional further appreciates that, with reference to the unit of each example of the embodiments described herein description
And algorithm steps, can be realized with electronic hardware, computer software or the combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software, generally describes the composition and step of each example according to function in the above description.These
Function is performed with hardware or software mode actually, depending on the application-specific and design constraint of technical scheme.Specialty
Technical staff can realize described function to each specific application using distinct methods, but this realization should not
Think beyond the scope of this invention.
The step of method or algorithm for being described with reference to the embodiments described herein, directly can be held with hardware, processor
Capable software module, or the two combination is implemented.Software module can be placed in random access memory (RAM), internal memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In field in known any other form of storage medium.
Dormitory safety monitoring system provided by the present invention and method are described in detail above.It is used herein
Specific case is set forth to principle of the invention and implementation method, and the explanation of above example is only intended to help and understands this
The method and its core concept of invention.It should be pointed out that for those skilled in the art, not departing from this hair
On the premise of bright principle, some improvement and modification can also be carried out to the present invention, these are improved and modification also falls into power of the present invention
In the protection domain that profit is required.
Claims (10)
1. a kind of dormitory safety monitoring system, it is characterised in that including：
White light camera, infrared camera, rotatable mechanical head framework, central processing unit and image processor；
Wherein, the white light camera and the infrared camera may be contained within the mechanical head framework, the white light
Camera is used for when current environment luminous intensity is higher than preset first threshold value, the video image in collection dormitory monitor area；Institute
Infrared camera is stated for when current environment luminous intensity is not higher than the preset first threshold value, gathering in dormitory monitor area
Video image；
The central processing unit is connected with the white light camera and the infrared camera, for regarding described in collecting
Frequency image is pre-processed, and is input into described image processor；
Described image processor is connected with the central processing unit, to be identified right in video image after pretreatment for extracting
The facial image feature of elephant, and the facial image feature is matched with the standard form for prestoring, it is described to judge
Whether object to be identified is dormitory member.
2. dormitory safety monitoring system as claimed in claim 1, it is characterised in that also include：
The infrared light-feeding instrument being connected with the infrared camera, for the current environment residing for the infrared camera ought to be detected
When luminous intensity is less than default Second Threshold, light filling operation is carried out.
3. dormitory safety monitoring system as claimed in claim 1 or 2, it is characterised in that the white light camera and described
Infrared camera is autozoom camera, for when described image processor judge the object to be identified as non-dormitory into
During member, object to be identified described in auto-focusing, and behavior to the object to be identified is tracked monitoring.
4. dormitory safety monitoring system as claimed in claim 3, it is characterised in that also include：
The warning device being connected with described image processor, for judging the object to be identified as non-when described image processor
During dormitory member, warning reminding is carried out in the form of voice and/or electric bell.
5. a kind of dormitory method for safety monitoring, it is characterised in that including：
Monitoring current environment luminous intensity, when the current environment luminous intensity is higher than preset first threshold value, using white light camera
Video image in collection dormitory monitor area；When the current environment luminous intensity is not higher than the preset first threshold value, adopt
The video image in dormitory monitor area is gathered with infrared camera；
The video image to collecting is pre-processed；
Extract the facial image feature of object to be identified in video image after pretreatment, and by the facial image feature with
The standard form for prestoring is matched, to judge whether the object to be identified is dormitory member.
6. dormitory method for safety monitoring as claimed in claim 5, it is characterised in that it is described by the facial image feature with it is pre-
The standard form for first storing is matched, to judge whether the object to be identified is that dormitory member includes：
When the ambient light intensity is not higher than the preset first threshold value, monitored using dynamic fluid recognition methods identification dormitory
Moving target in region；
After facial image is recognized, facial zone is divided into many sub-regions；
Characteristic point and Corner Detection are carried out to each facial subregion using image processor, the similar of double attribute mathematical models is calculated
Degree；
When the similarity after matching exceedes default similarity threshold, judge that the object to be identified is dormitory member；Otherwise sentence
The fixed object to be identified is non-dormitory member.
7. dormitory method for safety monitoring as claimed in claim 6, it is characterised in that facial zone is divided into many height described
Also include before region：
Facial image to recognizing carries out horizontal motion correction.
8. the dormitory method for safety monitoring as described in any one of claim 5 to 7, it is characterised in that also include：
When the object to be identified is judged as non-dormitory member, object to be identified described in auto-focusing, to described to be identified right
The behavior of elephant is tracked monitoring.
9. dormitory method for safety monitoring as claimed in claim 8, it is characterised in that also include：
When the object to be identified is judged as non-dormitory member, warning reminding is carried out in the form of voice and/or electric bell.
10. dormitory method for safety monitoring as claimed in claim 9, it is characterised in that also include：
When the current environment luminous intensity is detected less than default Second Threshold, light filling operation is carried out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710287122.7A CN106897716A (en) | 2017-04-27 | 2017-04-27 | A kind of dormitory safety monitoring system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710287122.7A CN106897716A (en) | 2017-04-27 | 2017-04-27 | A kind of dormitory safety monitoring system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106897716A true CN106897716A (en) | 2017-06-27 |
Family
ID=59197555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710287122.7A Pending CN106897716A (en) | 2017-04-27 | 2017-04-27 | A kind of dormitory safety monitoring system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106897716A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107343152A (en) * | 2017-08-25 | 2017-11-10 | 无锡北斗星通信息科技有限公司 | Real-time passenger image data correction system |
| CN107506752A (en) * | 2017-09-18 | 2017-12-22 | 艾普柯微电子（上海）有限公司 | Face identification device and method |
| CN107748860A (en) * | 2017-09-01 | 2018-03-02 | 中国科学院深圳先进技术研究院 | Method for tracking target, device, unmanned plane and the storage medium of unmanned plane |
| CN108388878A (en) * | 2018-03-15 | 2018-08-10 | 百度在线网络技术（北京）有限公司 | The method and apparatus of face for identification |
| CN108597168A (en) * | 2018-04-24 | 2018-09-28 | 广东美的制冷设备有限公司 | Security alarm method, apparatus based on optical filter and household appliance |
| CN109035120A (en) * | 2018-10-30 | 2018-12-18 | 深圳市海能通信股份有限公司 | A kind of big data security-protection management system |
| CN109145883A (en) * | 2018-10-10 | 2019-01-04 | 百度在线网络技术（北京）有限公司 | Method for safety monitoring, device, terminal and computer readable storage medium |
| CN109255867A (en) * | 2018-08-24 | 2019-01-22 | 星络科技有限公司 | Community's access control management method, device and computer storage medium |
| CN109598827A (en) * | 2018-09-25 | 2019-04-09 | 深圳神目信息技术有限公司 | A kind of face welcome hybrid system and its working method |
| CN109829888A (en) * | 2018-12-27 | 2019-05-31 | 北京林业大学 | Wild animal monitoring analysis system and method based on depth convolutional neural networks |
| CN110430393A (en) * | 2019-07-13 | 2019-11-08 | 恒大智慧科技有限公司 | A kind of monitoring method and system |
| CN111144182A (en) * | 2018-11-06 | 2020-05-12 | 天地融科技股份有限公司 | Method and system for detecting face risk in video |
| CN111144181A (en) * | 2018-11-06 | 2020-05-12 | 天地融科技股份有限公司 | Risk detection method, device and system based on background collaboration |
| CN111191498A (en) * | 2019-11-07 | 2020-05-22 | 腾讯科技（深圳）有限公司 | Behavior recognition method and related product |
| CN111814561A (en) * | 2020-06-11 | 2020-10-23 | 浙江大华技术股份有限公司 | Face recognition method, face recognition equipment and access control system |
| CN112019754A (en) * | 2020-09-16 | 2020-12-01 | 睿住科技有限公司 | Face recognition monitoring method and system |
| CN112669551A (en) * | 2020-12-23 | 2021-04-16 | 深圳砺剑天眼科技有限公司 | Object monitoring system and method based on infrared camera recognition |
| CN112731375A (en) * | 2020-12-11 | 2021-04-30 | 北京龙飞船科技有限公司 | Area detector capable of automatically adjusting sensitivity according to environmental change and method thereof |
| CN112887659A (en) * | 2021-01-29 | 2021-06-01 | 深圳前海微众银行股份有限公司 | Conference recording method, device, equipment and storage medium |
| CN113079350A (en) * | 2021-03-22 | 2021-07-06 | 昆仑数智科技有限责任公司 | Camera device and production monitored control system |
| CN113496209A (en) * | 2021-06-09 | 2021-10-12 | 湖南中惠旅智能科技有限责任公司 | Data processing method and system of scenic spot gate machine based on face recognition |
| CN115294680A (en) * | 2022-08-16 | 2022-11-04 | 杭州萤石软件有限公司 | Method and equipment for realizing personal safety protection function |
| CN115604572A (en) * | 2022-09-02 | 2023-01-13 | 荣耀终端有限公司（Cn） | Image acquisition method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260961A (en) * | 2015-10-15 | 2016-01-20 | 于璐菁 | Dormitory management system and management method thereof |
| CN105550680A (en) * | 2016-03-02 | 2016-05-04 | 北京盈想东方科技发展有限公司 | Beidou image recognition mobile monitoring terminal and monitoring method thereof |
| CN106200904A (en) * | 2016-06-27 | 2016-12-07 | 乐视控股（北京）有限公司 | A kind of gesture identifying device, electronic equipment and gesture identification method |
| CN206039671U (en) * | 2016-08-04 | 2017-03-22 | 合肥师范学院 | Face identification electronic lock |
-
2017
- 2017-04-27 CN CN201710287122.7A patent/CN106897716A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105260961A (en) * | 2015-10-15 | 2016-01-20 | 于璐菁 | Dormitory management system and management method thereof |
| CN105550680A (en) * | 2016-03-02 | 2016-05-04 | 北京盈想东方科技发展有限公司 | Beidou image recognition mobile monitoring terminal and monitoring method thereof |
| CN106200904A (en) * | 2016-06-27 | 2016-12-07 | 乐视控股（北京）有限公司 | A kind of gesture identifying device, electronic equipment and gesture identification method |
| CN206039671U (en) * | 2016-08-04 | 2017-03-22 | 合肥师范学院 | Face identification electronic lock |
Non-Patent Citations (1)
| Title |
|---|
| 王珏等编著: "《物业设备维护与管理》", 28 February 2010, 《东北财经大学出版社》 * |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107343152B (en) * | 2017-08-25 | 2018-03-20 | 杨丽 | Real-time passenger image data correction system |
| CN107343152A (en) * | 2017-08-25 | 2017-11-10 | 无锡北斗星通信息科技有限公司 | Real-time passenger image data correction system |
| CN107748860A (en) * | 2017-09-01 | 2018-03-02 | 中国科学院深圳先进技术研究院 | Method for tracking target, device, unmanned plane and the storage medium of unmanned plane |
| CN107506752A (en) * | 2017-09-18 | 2017-12-22 | 艾普柯微电子（上海）有限公司 | Face identification device and method |
| CN108388878A (en) * | 2018-03-15 | 2018-08-10 | 百度在线网络技术（北京）有限公司 | The method and apparatus of face for identification |
| CN108597168A (en) * | 2018-04-24 | 2018-09-28 | 广东美的制冷设备有限公司 | Security alarm method, apparatus based on optical filter and household appliance |
| CN109255867A (en) * | 2018-08-24 | 2019-01-22 | 星络科技有限公司 | Community's access control management method, device and computer storage medium |
| CN109598827A (en) * | 2018-09-25 | 2019-04-09 | 深圳神目信息技术有限公司 | A kind of face welcome hybrid system and its working method |
| CN109145883B (en) * | 2018-10-10 | 2019-10-22 | 百度在线网络技术（北京）有限公司 | Method for safety monitoring, device, terminal and computer readable storage medium |
| CN109145883A (en) * | 2018-10-10 | 2019-01-04 | 百度在线网络技术（北京）有限公司 | Method for safety monitoring, device, terminal and computer readable storage medium |
| CN109035120A (en) * | 2018-10-30 | 2018-12-18 | 深圳市海能通信股份有限公司 | A kind of big data security-protection management system |
| CN111144181A (en) * | 2018-11-06 | 2020-05-12 | 天地融科技股份有限公司 | Risk detection method, device and system based on background collaboration |
| CN111144182B (en) * | 2018-11-06 | 2023-04-07 | 天地融科技股份有限公司 | Method and system for detecting face risk in video |
| CN111144182A (en) * | 2018-11-06 | 2020-05-12 | 天地融科技股份有限公司 | Method and system for detecting face risk in video |
| CN109829888A (en) * | 2018-12-27 | 2019-05-31 | 北京林业大学 | Wild animal monitoring analysis system and method based on depth convolutional neural networks |
| CN110430393A (en) * | 2019-07-13 | 2019-11-08 | 恒大智慧科技有限公司 | A kind of monitoring method and system |
| CN110430393B (en) * | 2019-07-13 | 2021-06-04 | 星络智能科技有限公司 | Monitoring method and system |
| CN111191498A (en) * | 2019-11-07 | 2020-05-22 | 腾讯科技（深圳）有限公司 | Behavior recognition method and related product |
| CN111814561A (en) * | 2020-06-11 | 2020-10-23 | 浙江大华技术股份有限公司 | Face recognition method, face recognition equipment and access control system |
| CN112019754A (en) * | 2020-09-16 | 2020-12-01 | 睿住科技有限公司 | Face recognition monitoring method and system |
| CN112731375A (en) * | 2020-12-11 | 2021-04-30 | 北京龙飞船科技有限公司 | Area detector capable of automatically adjusting sensitivity according to environmental change and method thereof |
| CN112669551A (en) * | 2020-12-23 | 2021-04-16 | 深圳砺剑天眼科技有限公司 | Object monitoring system and method based on infrared camera recognition |
| CN112887659A (en) * | 2021-01-29 | 2021-06-01 | 深圳前海微众银行股份有限公司 | Conference recording method, device, equipment and storage medium |
| CN113079350A (en) * | 2021-03-22 | 2021-07-06 | 昆仑数智科技有限责任公司 | Camera device and production monitored control system |
| CN113496209A (en) * | 2021-06-09 | 2021-10-12 | 湖南中惠旅智能科技有限责任公司 | Data processing method and system of scenic spot gate machine based on face recognition |
| CN113496209B (en) * | 2021-06-09 | 2024-03-29 | 湖南中惠旅智能科技有限责任公司 | Data processing method and system of scenic spot gate machine based on face recognition |
| CN115294680A (en) * | 2022-08-16 | 2022-11-04 | 杭州萤石软件有限公司 | Method and equipment for realizing personal safety protection function |
| CN115604572A (en) * | 2022-09-02 | 2023-01-13 | 荣耀终端有限公司（Cn） | Image acquisition method and device |
| CN115604572B (en) * | 2022-09-02 | 2023-12-01 | 荣耀终端有限公司 | Image acquisition method, electronic device and computer readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106897716A (en) | A kind of dormitory safety monitoring system and method | |
| KR101970442B1 (en) | Illegal parking enforcement system Using Fast R-CNN based on Vehicle detection | |
| CN107423690A (en) | A kind of face identification method and device | |
| CN109918971B (en) | Method and device for detecting number of people in monitoring video | |
| CN111325150A (en) | Dressing detection device, method and storage medium | |
| CN106156688A (en) | A kind of dynamic human face recognition methods and system | |
| Li et al. | Face detection in complex background based on skin color features and improved AdaBoost algorithms | |
| CN103605969A (en) | Method and device for face inputting | |
| CN110728252B (en) | Face detection method applied to regional personnel motion trail monitoring | |
| Amit et al. | Efficient focusing and face detection | |
| CN112287827A (en) | Complex environment pedestrian mask wearing detection method and system based on intelligent lamp pole | |
| Chen et al. | Outdoor shadow estimating using multiclass geometric decomposition based on BLS | |
| CN111553214B (en) | Method and system for detecting smoking behavior of driver | |
| Saleem et al. | Face recognition using facial features | |
| CN106203338B (en) | Human eye state method for quickly identifying based on net region segmentation and threshold adaptive | |
| Rujikietgumjorn et al. | Real-time hog-based pedestrian detection in thermal images for an embedded system | |
| Rezaei et al. | Adaptive Haar-like classifier for eye status detection under non-ideal lighting conditions | |
| CN109543521A (en) | The In vivo detection and face identification method that main side view combines | |
| CN115331453A (en) | Non-contact intelligent control method for vehicle entering | |
| Wang et al. | Bus-crowdedness estimation by shallow convolutional neural network | |
| Yi et al. | Face detection method based on skin color segmentation and facial component localization | |
| CN106599847A (en) | Human body detection method based on depth information | |
| Archana et al. | Face recognition using Deep Neural Network and Support Based Anchor Boxes Methods in video frames | |
| KR100729265B1 (en) | A face detection method using difference image and color information | |
| Kubanek | Automatic methods for determining the characteristic points in face image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| 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: 20170627 |