CN106372576A - Deep learning-based intelligent indoor intrusion detection method and system - Google Patents

Deep learning-based intelligent indoor intrusion detection method and system Download PDF

Info

Publication number
CN106372576A
CN106372576A CN201610705858.7A CN201610705858A CN106372576A CN 106372576 A CN106372576 A CN 106372576A CN 201610705858 A CN201610705858 A CN 201610705858A CN 106372576 A CN106372576 A CN 106372576A
Authority
CN
China
Prior art keywords
image
user
images
deep learning
network model
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
Application number
CN201610705858.7A
Other languages
Chinese (zh)
Inventor
胡婧
覃婷婷
成孝刚
邵文泽
成云
李德志
李海波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Post and Telecommunication University
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing Post and Telecommunication University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing Post and Telecommunication University filed Critical Nanjing Post and Telecommunication University
Priority to CN201610705858.7A priority Critical patent/CN106372576A/en
Publication of CN106372576A publication Critical patent/CN106372576A/en
Pending legal-status Critical Current

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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural 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
    • G06V40/10—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
    • 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)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Computing Systems (AREA)
  • Software Systems (AREA)
  • Molecular Biology (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Signal Processing (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a deep learning-based intelligent indoor intrusion detection method and system. The method comprises the following steps of establishing a BP neural network model; obtaining difference images between adjacent frames in a monitoring video picture by utilizing a frame difference algorithm; carrying out binarization processing on the obtained difference images and extracting change foreground area images, under a static background, in the processed images; detecting and identifying whether human shapes exist in the extracted change foreground images or not; when the human shapes exist, detecting and extracting face area images from the change foreground areas; identifying whether the extracted face area images are user images or not; and when the face area images are not the user images, determining the face area images as non-user intrusion and sending alarm signals to the users. The method and system are strong in ability of resisting the interference of other moving objects and low in mis-judgement rate, and can be used for carrying out massive video data analysis and correctly carrying out intrusion detection and identification.

Description

A kind of Intelligent indoor intrusion detection method based on deep learning and system
Technical field
The present invention relates to a kind of Intelligent indoor intrusion detection method based on deep learning and system, belong to video monitoring Technical field.
Background technology
With economic and science and technology development, people are for the attention degree also more and more higher of indoor security, intelligent room Interior intruding detection system becomes trend, then the intelligent, accuracy for intrusion target detection and identification and real-time Research also become very meaningful.
In traditional intelligent control method, moving object whether is had to intrude into it using photographic head and Infrared Detectorss perception Sensing range, if invader is detected, video information is sent to mobile terminal.
But traditional monitoring method existing defects, specifically, its method cannot differentiate whether invader is user, resists other Moving object interference performance is little, and False Rate is high, intelligent low, does not possess accurately detection identification function；It is due to processor Limit it is impossible to carry out substantial amounts of video data analysis, real-time is low.Accordingly, there exist to user and invader, moving object with enter The person of invading can not accurately detect the problem of identification.
Content of the invention
The technical problem to be solved is to overcome the deficiencies in the prior art, provide a kind of based on deep learning Intelligent indoor intrusion detection method and system, solve to user and invader, moving object and invader can not accurately detect knowledge Other problem.
The present invention specifically employs the following technical solutions solution above-mentioned technical problem:
A kind of Intelligent indoor intrusion detection method based on deep learning, comprises the following steps:
Set up bp neural network model, and according to the training data that input comprises user images, bp neural network model is instructed Practice；
The difference image of adjacent interframe is obtained using inter-frame difference algorithm from monitor video picture；
Binary conversion treatment is carried out to the difference image of acquired adjacent interframe, and extracts the static back of the body in image after treatment Prospects For Changes region under scape；
To the Prospects For Changes image detection extracted and identify that it whether there is using the bp neural network model set up Humanoid；When identification has humanoid, detection and extraction from described Prospects For Changes image obtains human face region image；
Using the bp neural network model set up, the human face region image detection extracted and identification are determined whether User images；And when identification is judged as non-user image, is defined as non-user invasion and sends alarm signal to user.
Further, as a preferred technical solution of the present invention, in methods described, human face region image is by following step Rapid acquisition:
Set up based on colored complexion model, and normalization colour of skin similarity；
Using the complexion model set up to Prospects For Changes image detection, obtain normalized colour of skin similarity graph picture；
Set colour of skin similarity threshold, binary conversion treatment is carried out to acquired normalized colour of skin similarity graph picture, obtain Area of skin color；
According to the scope setting, area of skin color is chosen, obtain human face region image.
Further, as a preferred technical solution of the present invention: utilize prp conjugate gradient algorithms pair in methods described Bp neural network model is trained.
Further, as a preferred technical solution of the present invention: in methods described, training data at least includes possessing User's whole body images of different attitudes and user's facial image.
Further, as a preferred technical solution of the present invention: in methods described training data also include non-humanoid Image.Further, as a preferred technical solution of the present invention: in methods described when identification judges not have humanoid, Reacquire monitor video picture.
Further, as a preferred technical solution of the present invention: be judged as user images in identification in methods described When, it is defined as user and monitor video picture is reacquired according to setting time.
The present invention also proposes a kind of Intelligent indoor intruding detection system based on deep learning, comprising:
Video monitoring module, for domestic environment is entered with Mobile state monitoring, obtains monitor video picture；
Image capture module, for obtaining the difference diagram of adjacent interframe from monitor video picture using inter-frame difference algorithm Picture；
Image zooming-out module, for carrying out binary conversion treatment to the difference image of acquired adjacent interframe, and extracts through place Prospects For Changes area image under static background in image after reason；
Discriminatory analysiss module, including modeling unit, recognition unit, area extracting unit and output unit；Wherein, described build Form unit, is used for setting up bp neural network model, and the training data comprising user images according to input is to bp neutral net mould Type training；Described recognition unit, for being detected the Prospects For Changes being extracted image input bp neural network model and being known Do not judge whether humanoid；Described area extracting unit is used for when identifying that judgement has humanoid, from described Prospects For Changes figure In picture, detection and extraction obtain human face region image；Described recognition unit is additionally operable to the human face region image extracting input bp god Detected through network model and identification is determined whether user images；Described output unit, for being judged as non-use in identification It is defined as non-user invasion during the image of family, and generate and output control signal；
Transmitting element, for sending invasion prompting data according to control signal to user.
Further, as a preferred technical solution of the present invention: the invasion prompting packet sending in this unit Include invader's image.
Further, as a preferred technical solution of the present invention: also include alarm module, described alarm module is used for Sound and light alarm is produced according to control signal.
The present invention adopts technique scheme, can produce following technique effect:
Intelligent indoor intrusion detection method based on deep learning provided by the present invention and system, first to monitor video Moving target in picture is detected, so not carrying out subsequent treatment for the constant video image of scene, saves server Expense；And add in systems to humanoid detection, it is to avoid house pet, the interference of the inhuman factor such as sweeping robot；Because right The identification of user is eventually transformed into the identification to user's face, so the change of user's dress ornament hair style has no effect on the correct knowledge to it , there is no training data needed for good universality, neutral net few.And, image is gathered using infrared camera, can not be subject to Illumination limits；Therefore design more humane, other moving object interference performances anti-are strong, False Rate is low, it is possible to use high performance Processing mode carries out substantial amounts of video data analysis, possesses high efficiency and real-time, can transport exactly to user and invader Animal body and invader accurately detect identification.
The method of the present invention and system possess following advantage:
(1) the Intelligent indoor intrusion detection method based on deep learning provided by the present invention and system, using based on prp The bp neutral net of conjugate gradient algorithms, improves network training speed it is ensured that restraining correctness.
(2) present invention incorporates inter-frame difference algorithm and skin cluster model, to target area grading extraction, calculate letter Single it is ensured that real-time.
(3) extract the target area input neural network detecting, be not required to identify complex background and other interference factors, protect The correctness of card classification.
(4) present invention is based on image processing techniquess, can expeditiously complete to identify, can save power consumption.Can effectively solve Certainly to user and invader, moving object and invader can not accurately detect the problem of identification.
Brief description
Fig. 1 is the schematic flow sheet of the Intelligent indoor intrusion detection method based on deep learning of the present invention.
Fig. 2 is the bp Artificial Neural Network Structures schematic diagram of the present invention.
Fig. 3 is the principle schematic of the Intelligent indoor intruding detection system based on deep learning of the present invention.
Specific embodiment
With reference to Figure of description, embodiments of the present invention are described.
As shown in figure 1, the present invention devises a kind of Intelligent indoor intrusion detection method based on deep learning, including following Step:
Step 1, set up bp neural network model, and the training data comprising user images according to input is to bp neutral net Model training.In methods described, training data to be inputted using this model for the first time, bp neural network model is trained, its Step includes initializing weights and threshold value, successively adjusts weights and threshold value respectively using prp conjugate gradient algorithms, iteration is to maximum Iterationses.Its process is specific as follows:
To the bp neural network model input training data using prp conjugate gradient algorithms, described training data at least wraps Include the user's whole body images possessing different attitudes and user's facial image it is preferable that can also include non-humanoid image such as house pet, Intellective dust collector etc..
As shown in Fig. 2 f is a class training data, neutral net is output as y=ψ to the iterative process of bp neural network model (wixi- θ), wherein ψ represents excitation function, can use sigma function；xi, xjRepresent i-th layer and the input of jth layer neuron respectively Vector；wiRepresent the weight vector connecting i-th layer and next layer, wjRepresent the weight vector connecting jth layer and next layer；θ is threshold Value.Using i in Fig. 2, referring to all hidden layers in neutral net, hidden layers numbers determine j two-layer hidden layer according to practical situation, because respectively Layer alternative manner is identical, so hereafter only referring to the input vector x of each layerlWith weight vector wl, have
Prp conjugate gradient algorithms convergence weights and threshold value is adopted in the present invention.Then object function is:
l(wl, θ) and=wlxl-θ
And, set the object function error gradient that kth time iteration obtains as ek, be can get based on prp conjugate gradient algorithms CoefficientThus can determine that the direction of search: pk=-ek+βkpk-1.Step-size in search α is obtained by linear searchk, lead to Cross that conjugate gradient algorithms obtain weight vector and the iterative difference of input vector is as follows:
wl(k+1)=wl(k)+αkpk, k=0,1 ...；
θ (k+1)=θ (k)+αkpk, k=0,1 ....
W in above-mentioned formulalK () represents kth time iteration wlThe weight vector iterative value obtaining, θ (k) represents that kth time iteration θ obtains To threshold value iterative value.
And, the bp neural network learning process of the present invention adopted prp conjugate gradient algorithms is as described below:
First, input: a class training image f={ f1, f2..., fn, wherein f can comprise the people possessing different attitudes Shape whole body images or user's facial image.
Subsequently into iterative process, as follows:
1. initialize；Initialize each layer neuron weight vector wl(0), threshold θ (0).
2. adjust weights；Fixed threshold θ, successively adjusts each layer weight vector w using weight vector is iterativel, until convergence.
3. adjust threshold value；Fixing weight wl, using threshold value iterative adjustment threshold θ, until convergence.
4. iteration, repetitive process 2,3, until it reaches maximum iteration time.
Possess the humanoid whole body images of different attitudes to the bp neural network model input using prp conjugate gradient algorithms, Trained study model can be made to obtain Human detection device function, can recognize that humanoid and non-humanoid region；And, input user Facial image, makes model obtain face identification functions, the invader of recognizable user and non-user with trained study.
Thus obtained bp neural network model is capable of identify that user and invader, possesses the humanoid mankind and does not possess people The moving object of shape.
Step 2, obtain the difference image of adjacent interframe from monitor video picture using inter-frame difference algorithm.It is from prison Control video pictures capture multiple image, specifically includes following steps:
If pth frame and pth+1 frame inputted video image are respectively fp(i, j) and fp+1(i, j), wherein (i, j) represent pixel Point, by corresponding for two field pictures pixel difference:
dp(i, j)=| fp+1(i, j)-fp(i, j) |
Then can obtain difference image:
Step 3, the difference image to acquired adjacent interframe carry out binary conversion treatment, and it is black for obtaining static background, change Foreground area is white bianry image, therefore can by Prospects For Changes extracted region out, to be provided in recognition detection process.And it is right It is not required to carry out subsequent treatment to background area in the constant video image of stationary background, server overhead can be saved.
The Prospects For Changes image zooming-out of prospect in described image, particularly as follows:
Given threshold t, obtains the bianry image of Prospects For Changes image:
h p ( i , j ) = 1 , d p ( i , j ) &greaterequal; t 0 , d p ( i , j ) < t
Step 4, using the bp neural network model set up to the Prospects For Changes image detection extracted and identification judge With the presence or absence of humanoid.Specifically include following steps:
After the above-mentioned image-region that moving target in static background is located extracts, at the image of this extraction Reason and cutting, to mate neural network model input picture characteristic.Then it is inputted in bp neural network model, using bp god Through network model image is carried out humanoid with non-humanoid identification, this step is using to humanoid detection, it is to avoid house pet, sweeps the floor The interference of the inhuman factor such as robot.If bp neural network model exports y=1, there is humanoid region, continue in Prospects For Changes image Continue and human face detection and tracing operation is carried out to target area, that is, enter step 5；Otherwise, bp neural network model output y=0, table Bright do not have non-humanoid region, does not carry out following step 5, continues monitoring.I.e. when identification judges not have humanoid, show inhuman Body can reacquire monitor video picture and repeat step 3 to 4 with return to step 2, carry out continuation and regard in active state, this method Frequency monitors and identifies, until it is humanoid to identify that judgement exists.
Export y=1 in bp neural network model, when identification judgement has humanoid, detect from described Prospects For Changes image Obtain human face region image with extracting.Described human face region image acquisition procedures are specific as follows:
First, using the complexion model based on y cb cr color system, pixel (i, j) place chroma blue component can be obtained Value cbWith red chrominance component value cr；If the chromatic value x of pixelij=(cb, cr)t, then the colour of skin similarity of pixel (i, j) For:
p(cb, cr)=exp [- 0.5 (xij-m)tc-1(xij-m)]
Wherein, m is xijAverage, c is xijVariance.
Again by colour of skin similarity max p (c maximum in regionb, cr), to p (cb, cr) it is normalized:
Normalized colour of skin similarity graph can be obtained as f (i, j)=p ' (cb, cr).Setting colour of skin similarity adaptive thresholding Value q, by following formula acquisition bianry image:
f t ( i , j ) = 1 , f ( i , j ) &greaterequal; q 0 , f ( i , j ) < q
Area of skin color maximum height c then can be obtainedmaxWith maximum width lmax, face is obtained based on Face geometric eigenvector Approximate region, step is as follows:
(1) area of skin color in vertical direction is respectively widened up and down 0.5*cmax, make this point ft(x, y) is 1, as face Region.
(2) area of skin color in horizontal direction is respectively widened up and down 0.5*lmax, make this point ft(x, y) is 1, as face Region.
Thus, it is possible to obtain human face region image, then it is identified in input following step 5.
Step 5, using the bp neural network model set up to the human face region image detection extracted and identification judge Whether it is user images.Specifically include following steps:
By the human face region image zooming-out of step 4 out after, the image procossing to this extraction and cutting, to mate nerve Network model's input picture characteristic.Then it is inputted in bp neural network model, recognition of face is carried out to it.
If bp neural network model exports y=0, this face is non-user image, shows there is invader；Sentence in identification When breaking as non-user image, it is defined as non-user invasion and sends alarm signal to user.
Otherwise, bp neural network model output y=1, is identified as user images.It is judged as user's facial image in identification When, it is defined as user and monitor video picture can be reacquired according to setting time, be i.e. return to step 2 to 5, to complete to continue Video monitoring and identification process.
On this basis, the present invention also proposes a kind of Intelligent indoor intruding detection system based on deep learning, this system Image acquisition, extraction and recognition detection can be carried out using above-mentioned detection method.This system just can be divided into according to user setup Often mode of operation and park mode.When there being visitor's visiting, user can open park mode, can save power consumption, and design is more humane. And when user leaves, activation system can be selected, open normal mode of operation so that it is monitored to household.
Specifically, described system includes:
Video monitoring module, for domestic environment is entered with Mobile state monitoring, obtains monitor video picture；
Image capture module, for obtaining the difference diagram of adjacent interframe from monitor video picture using inter-frame difference algorithm Picture；
Image zooming-out module, for carrying out binary conversion treatment to the difference image of acquired adjacent interframe, and to treated Image zooming-out afterwards obtains the Prospects For Changes image under static background；
Discriminatory analysiss module, including modeling unit, recognition unit, area extracting unit and output unit；Wherein, described build Form unit, is used for setting up bp neural network model, and the training data comprising user images according to input is to bp neutral net mould Type training；Described recognition unit, for being detected the Prospects For Changes being extracted image input bp neural network model and being known Do not judge whether humanoid；Described area extracting unit is used for when identifying that judgement has humanoid, from described Prospects For Changes figure In picture, detection and extraction obtain human face region image；Described recognition unit is additionally operable to the human face region image extracting input bp god Detected through network model and identification is determined whether user images；Described output unit, for being judged as non-use in identification It is defined as non-user invasion during the image of family, and generate and output control signal；
Transmitting element, for sending invasion prompting data according to control signal to user.
Described system, starts under normal mode of operation, video monitoring module can be under the different situations of day and night Household is monitored, preferably multiframe input picture can be collected by infrared monitoring camera electronic equipment, illumination can not be subject to Limit, to obtain monitor video picture.
And, its principle of described system is as shown in figure 3, modeling unit in discriminatory analysiss module, after setting up model, Training data to be inputted using model for the first time, bp neural network model is trained, its step includes initializing weights and threshold Value, successively adjusts weights and threshold value respectively using prp conjugate gradient algorithms, iteration is to maximum iteration time.And recognition unit, area Described in the engineering process of domain extraction unit and output unit method all described above.
In order to monitoring to household is better achieved, described system can also include alarm module, and described alarm module is used According to control signal generation sound and light alarm.If system is under normal mode of operation, when being defined as non-user invasion, to User sends invasion prompting data and by alarm module, sound and light alarm occurs；Wherein, invasion prompting data can include extracting The invader humanoid figure picture obtaining or invader's facial image, or both of which transmission.
If system is in park mode, transmitting element and alarm module all receive control signal, but do not send invasion and carry Registration evidence and generation are reported to the police.
To sum up, the Intelligent indoor intrusion detection method based on deep learning provided by the present invention and system, using being based on The bp neutral net of prp conjugate gradient algorithms, improves network training speed it is ensured that restraining correctness and ensure that real-time.Carry Take the target area input neural network detecting, be not required to identify complex background and other interference factors it is ensured that classifies is correct Property.Therefore can expeditiously complete to identify, power consumption can be saved.Can with effectively solving to user and invader, moving object with Invader can not accurately detect the problem of identification.
Above in conjunction with accompanying drawing, embodiments of the present invention are explained in detail, but the present invention is not limited to above-mentioned enforcement Mode, in the ken that those of ordinary skill in the art possess, can also be on the premise of without departing from present inventive concept Make a variety of changes.

Claims (10)

1. a kind of Intelligent indoor intrusion detection method based on deep learning is it is characterised in that comprise the following steps:
Set up bp neural network model, and according to the training data that input comprises user images, bp neural network model is trained；
The difference image of adjacent interframe is obtained using inter-frame difference algorithm from monitor video picture；
Binary conversion treatment is carried out to above-mentioned difference image, the Prospects For Changes region in the image after extraction process；
To the Prospects For Changes image detection extracted and identify it with the presence or absence of humanoid using the bp neural network model set up； When identification has humanoid, detection and extraction from described Prospects For Changes image obtains human face region image；
Using the bp neural network model set up, user is determined whether to the human face region image detection extracted and identification Image；And when identification is judged as non-user image, is defined as non-user invasion and sends alarm signal to user.
2. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that methods described Middle human face region image is obtained by following steps:
Set up based on colored complexion model, and normalization colour of skin similarity；
Using the complexion model set up to Prospects For Changes image detection, obtain normalized colour of skin similarity graph picture；
Set colour of skin similarity threshold, acquired normalized colour of skin similarity graph picture is carried out with binary conversion treatment, obtain the colour of skin Region；
According to the scope setting, area of skin color is chosen, obtain human face region image.
3. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that: methods described Middle utilization prp conjugate gradient algorithms are trained to bp neural network model.
4. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that: methods described Middle training data at least includes user's whole body images and the user's facial image possessing different attitudes.
5. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that: methods described Middle training data also includes non-humanoid image.
6. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that: methods described In when identification judge do not have humanoid when, reacquire monitor video picture.
7. according to claim 1 the Intelligent indoor intrusion detection method based on deep learning it is characterised in that: methods described In when identification is judged as user images, be defined as user and according to setting time reacquire monitor video picture.
8. a kind of Intelligent indoor intruding detection system based on deep learning is it is characterised in that include:
Video monitoring module, for domestic environment is entered with Mobile state monitoring, obtains monitor video picture；
Image capture module, for obtaining the difference image of adjacent interframe from monitor video picture using inter-frame difference algorithm；
Image zooming-out module, for carrying out binary conversion treatment to the difference image of acquired adjacent interframe, and extracts after treatment Image in Prospects For Changes region；
Discriminatory analysiss module, including modeling unit, recognition unit, area extracting unit and output unit；Wherein, described modeling is single Unit, is used for setting up bp neural network model, and according to the training data that input comprises user images, bp neural network model is instructed Practice；Described recognition unit, for inputting the Prospects For Changes being extracted image, bp neural network model is detected and identification is sentenced Break with the presence or absence of humanoid；Described area extracting unit is used for when identifying that judgement has humanoid, from described Prospects For Changes image Detection and extraction obtain human face region image；Described recognition unit is additionally operable to the human face region image extracting input bp nerve net Network model is detected and identification determines whether user images；Described output unit, for being judged as non-user figure in identification As when be defined as non-user invasion, and generate and output control signal；
Transmitting element, for sending invasion prompting data according to control signal to user.
9. according to claim 8 the Intelligent indoor intruding detection system based on deep learning it is characterised in that: described transmission The invasion prompting data that unit sends includes invader's image.
10. according to claim 8 the Intelligent indoor intruding detection system based on deep learning it is characterised in that: also include Alarm module, described alarm module is used for producing sound and light alarm according to control signal.
CN201610705858.7A 2016-08-23 2016-08-23 Deep learning-based intelligent indoor intrusion detection method and system Pending CN106372576A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610705858.7A CN106372576A (en) 2016-08-23 2016-08-23 Deep learning-based intelligent indoor intrusion detection method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610705858.7A CN106372576A (en) 2016-08-23 2016-08-23 Deep learning-based intelligent indoor intrusion detection method and system

Publications (1)

Publication Number Publication Date
CN106372576A true CN106372576A (en) 2017-02-01

Family

ID=57878466

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610705858.7A Pending CN106372576A (en) 2016-08-23 2016-08-23 Deep learning-based intelligent indoor intrusion detection method and system

Country Status (1)

Country Link
CN (1) CN106372576A (en)

Cited By (36)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106919912A (en) * 2017-02-16 2017-07-04 北京小米移动软件有限公司 The method and device of Indoor Video
CN107316024A (en) * 2017-06-28 2017-11-03 北京博睿视科技有限责任公司 perimeter alarm algorithm based on deep learning
CN107368818A (en) * 2017-07-27 2017-11-21 北京小米移动软件有限公司 Meeting room state method to set up, device, system and storage medium
CN107766829A (en) * 2017-10-27 2018-03-06 浙江大华技术股份有限公司 A kind of method and apparatus of Articles detecting
CN107820619A (en) * 2017-09-21 2018-03-20 达闼科技（北京）有限公司 One kind classification interactive decision making method, interactive terminal and cloud server
CN107818651A (en) * 2017-10-27 2018-03-20 华润电力技术研究院有限公司 A kind of illegal cross-border warning method and device based on video monitoring
CN108229359A (en) * 2017-12-26 2018-06-29 大唐软件技术股份有限公司 A kind of face image processing process and device
CN108280953A (en) * 2018-03-27 2018-07-13 上海小蚁科技有限公司 Video detecting alarm method and device, storage medium, camera
CN108416797A (en) * 2018-02-27 2018-08-17 鲁东大学 A kind of method, equipment and the storage medium of detection Behavioral change
CN108520611A (en) * 2018-04-12 2018-09-11 上海小蚁科技有限公司 Video data monitoring method and device, storage medium, terminal
CN108828576A (en) * 2018-04-10 2018-11-16 上海摩软通讯技术有限公司 Indoor locating system and method
CN109410497A (en) * 2018-11-20 2019-03-01 江苏理工学院 A kind of monitoring of bridge opening space safety and alarm system based on deep learning
CN109409315A (en) * 2018-11-07 2019-03-01 浩云科技股份有限公司 A kind of ATM machine panel zone remnant object detection method and system
CN109614906A (en) * 2018-12-03 2019-04-12 北京工业大学 A kind of security system and security alarm method based on deep learning
CN109977826A (en) * 2019-03-15 2019-07-05 百度在线网络技术（北京）有限公司 The classification recognition methods of object and device
CN110097050A (en) * 2019-04-03 2019-08-06 平安科技（深圳）有限公司 Pedestrian detection method, device, computer equipment and storage medium
CN110427815A (en) * 2019-06-24 2019-11-08 特斯联（北京）科技有限公司 Realize the method for processing video frequency and device of the effective contents interception of gate inhibition
CN110738079A (en) * 2018-07-19 2020-01-31 杭州海康威视数字技术股份有限公司 Method and device for detecting abnormal number of front row personnel of motor vehicle and computer equipment
CN110794462A (en) * 2019-11-06 2020-02-14 广东博智林机器人有限公司 Building site safety monitoring system and monitoring method and device thereof
CN111062415A (en) * 2019-11-12 2020-04-24 中南大学 Target object image extraction method and system based on contrast difference and storage medium
CN111382694A (en) * 2020-03-06 2020-07-07 杭州宇泛智能科技有限公司 Face recognition method and device and electronic equipment
WO2020168960A1 (en) * 2019-02-19 2020-08-27 杭州海康威视数字技术股份有限公司 Video analysis method and apparatus
US10834365B2 (en) 2018-02-08 2020-11-10 Nortek Security & Control Llc Audio-visual monitoring using a virtual assistant
CN112004056A (en) * 2020-08-06 2020-11-27 武汉倍特威视系统有限公司 Intelligent video analysis method with strong anti-interference capability
CN112291535A (en) * 2020-12-24 2021-01-29 南京斯酷环境科技有限公司 Public safety monitoring system and method based on Internet of things
CN112380962A (en) * 2020-11-11 2021-02-19 成都摘果子科技有限公司 Animal image identification method and system based on deep learning
US10978050B2 (en) 2018-02-20 2021-04-13 Intellivision Technologies Corp. Audio type detection
CN112699757A (en) * 2020-12-24 2021-04-23 武汉龙发包装有限公司 Goods elevator safety monitoring method and system for transporting cartons
CN112733676A (en) * 2020-12-31 2021-04-30 青岛海纳云科技控股有限公司 Method for detecting and identifying garbage in elevator based on deep learning
CN112802052A (en) * 2021-01-19 2021-05-14 北京小米移动软件有限公司 Image recognition method and device, electronic equipment and storage medium
CN113055654A (en) * 2021-03-26 2021-06-29 太原师范学院 Method for lossy compression of video stream in edge device
CN113810837A (en) * 2020-06-16 2021-12-17 京东方科技集团股份有限公司 Synchronous sounding control method of display device and related equipment
US11295139B2 (en) 2018-02-19 2022-04-05 Intellivision Technologies Corp. Human presence detection in edge devices
US11615623B2 (en) 2018-02-19 2023-03-28 Nortek Security & Control Llc Object detection in edge devices for barrier operation and parcel delivery
CN115909215A (en) * 2022-12-09 2023-04-04 厦门农芯数字科技有限公司 Edge intrusion early warning method and system based on target detection
US11735018B2 (en) 2018-03-11 2023-08-22 Intellivision Technologies Corp. Security system with face recognition

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521578A (en) * 2011-12-19 2012-06-27 中山爱科数字科技股份有限公司 Method for detecting and identifying intrusion
CN103067460A (en) * 2012-12-14 2013-04-24 厦门天聪智能软件有限公司 Corrective biology identification long distance identity checking method towards judicial community
CN103679212A (en) * 2013-12-06 2014-03-26 无锡清华信息科学与技术国家实验室物联网技术中心 Method for detecting and counting personnel based on video image
CN104239851A (en) * 2014-07-25 2014-12-24 重庆科技学院 Intelligent cell inspection system based on behavior analysis and control method thereof
CN104539874A (en) * 2014-06-17 2015-04-22 武汉理工大学 Human body mixed monitoring system and method fusing pyroelectric sensing with cameras
CN105760835A (en) * 2016-02-17 2016-07-13 天津中科智能识别产业技术研究院有限公司 Gait segmentation and gait recognition integrated method based on deep learning
CN105788126A (en) * 2016-04-29 2016-07-20 浙江理工大学 Intelligent household monitoring system and control method thereof

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102521578A (en) * 2011-12-19 2012-06-27 中山爱科数字科技股份有限公司 Method for detecting and identifying intrusion
CN103067460A (en) * 2012-12-14 2013-04-24 厦门天聪智能软件有限公司 Corrective biology identification long distance identity checking method towards judicial community
CN103679212A (en) * 2013-12-06 2014-03-26 无锡清华信息科学与技术国家实验室物联网技术中心 Method for detecting and counting personnel based on video image
CN104539874A (en) * 2014-06-17 2015-04-22 武汉理工大学 Human body mixed monitoring system and method fusing pyroelectric sensing with cameras
CN104239851A (en) * 2014-07-25 2014-12-24 重庆科技学院 Intelligent cell inspection system based on behavior analysis and control method thereof
CN105760835A (en) * 2016-02-17 2016-07-13 天津中科智能识别产业技术研究院有限公司 Gait segmentation and gait recognition integrated method based on deep learning
CN105788126A (en) * 2016-04-29 2016-07-20 浙江理工大学 Intelligent household monitoring system and control method thereof

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
时升云: ""基于智能视频监控的人流量统计系统研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *
陈松: ""基于肤色的人脸检测与识别"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (42)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106919912A (en) * 2017-02-16 2017-07-04 北京小米移动软件有限公司 The method and device of Indoor Video
CN107316024A (en) * 2017-06-28 2017-11-03 北京博睿视科技有限责任公司 perimeter alarm algorithm based on deep learning
CN107368818A (en) * 2017-07-27 2017-11-21 北京小米移动软件有限公司 Meeting room state method to set up, device, system and storage medium
CN107820619B (en) * 2017-09-21 2019-12-10 达闼科技（北京）有限公司 hierarchical interaction decision-making method, interaction terminal and cloud server
CN107820619A (en) * 2017-09-21 2018-03-20 达闼科技（北京）有限公司 One kind classification interactive decision making method, interactive terminal and cloud server
WO2019056267A1 (en) * 2017-09-21 2019-03-28 达闼科技（北京）有限公司 Hierarchical interactive decision making method, interactive terminal, and cloud server
CN107818651A (en) * 2017-10-27 2018-03-20 华润电力技术研究院有限公司 A kind of illegal cross-border warning method and device based on video monitoring
CN107766829A (en) * 2017-10-27 2018-03-06 浙江大华技术股份有限公司 A kind of method and apparatus of Articles detecting
CN108229359A (en) * 2017-12-26 2018-06-29 大唐软件技术股份有限公司 A kind of face image processing process and device
US10834365B2 (en) 2018-02-08 2020-11-10 Nortek Security & Control Llc Audio-visual monitoring using a virtual assistant
US11295139B2 (en) 2018-02-19 2022-04-05 Intellivision Technologies Corp. Human presence detection in edge devices
US11615623B2 (en) 2018-02-19 2023-03-28 Nortek Security & Control Llc Object detection in edge devices for barrier operation and parcel delivery
US10978050B2 (en) 2018-02-20 2021-04-13 Intellivision Technologies Corp. Audio type detection
CN108416797A (en) * 2018-02-27 2018-08-17 鲁东大学 A kind of method, equipment and the storage medium of detection Behavioral change
US11735018B2 (en) 2018-03-11 2023-08-22 Intellivision Technologies Corp. Security system with face recognition
CN108280953A (en) * 2018-03-27 2018-07-13 上海小蚁科技有限公司 Video detecting alarm method and device, storage medium, camera
CN108828576A (en) * 2018-04-10 2018-11-16 上海摩软通讯技术有限公司 Indoor locating system and method
CN108520611A (en) * 2018-04-12 2018-09-11 上海小蚁科技有限公司 Video data monitoring method and device, storage medium, terminal
CN110738079A (en) * 2018-07-19 2020-01-31 杭州海康威视数字技术股份有限公司 Method and device for detecting abnormal number of front row personnel of motor vehicle and computer equipment
CN109409315B (en) * 2018-11-07 2022-01-11 浩云科技股份有限公司 Method and system for detecting remnants in panel area of ATM (automatic Teller machine)
CN109409315A (en) * 2018-11-07 2019-03-01 浩云科技股份有限公司 A kind of ATM machine panel zone remnant object detection method and system
CN109410497A (en) * 2018-11-20 2019-03-01 江苏理工学院 A kind of monitoring of bridge opening space safety and alarm system based on deep learning
CN109614906A (en) * 2018-12-03 2019-04-12 北京工业大学 A kind of security system and security alarm method based on deep learning
WO2020168960A1 (en) * 2019-02-19 2020-08-27 杭州海康威视数字技术股份有限公司 Video analysis method and apparatus
CN109977826A (en) * 2019-03-15 2019-07-05 百度在线网络技术（北京）有限公司 The classification recognition methods of object and device
CN110097050B (en) * 2019-04-03 2024-03-08 平安科技（深圳）有限公司 Pedestrian detection method, device, computer equipment and storage medium
CN110097050A (en) * 2019-04-03 2019-08-06 平安科技（深圳）有限公司 Pedestrian detection method, device, computer equipment and storage medium
CN110427815B (en) * 2019-06-24 2020-07-10 特斯联（北京）科技有限公司 Video processing method and device for realizing interception of effective contents of entrance guard
CN110427815A (en) * 2019-06-24 2019-11-08 特斯联（北京）科技有限公司 Realize the method for processing video frequency and device of the effective contents interception of gate inhibition
CN110794462A (en) * 2019-11-06 2020-02-14 广东博智林机器人有限公司 Building site safety monitoring system and monitoring method and device thereof
CN110794462B (en) * 2019-11-06 2021-08-03 广东博智林机器人有限公司 Building site safety monitoring system and monitoring method and device thereof
CN111062415A (en) * 2019-11-12 2020-04-24 中南大学 Target object image extraction method and system based on contrast difference and storage medium
CN111382694A (en) * 2020-03-06 2020-07-07 杭州宇泛智能科技有限公司 Face recognition method and device and electronic equipment
CN113810837A (en) * 2020-06-16 2021-12-17 京东方科技集团股份有限公司 Synchronous sounding control method of display device and related equipment
CN112004056A (en) * 2020-08-06 2020-11-27 武汉倍特威视系统有限公司 Intelligent video analysis method with strong anti-interference capability
CN112380962A (en) * 2020-11-11 2021-02-19 成都摘果子科技有限公司 Animal image identification method and system based on deep learning
CN112699757A (en) * 2020-12-24 2021-04-23 武汉龙发包装有限公司 Goods elevator safety monitoring method and system for transporting cartons
CN112291535A (en) * 2020-12-24 2021-01-29 南京斯酷环境科技有限公司 Public safety monitoring system and method based on Internet of things
CN112733676A (en) * 2020-12-31 2021-04-30 青岛海纳云科技控股有限公司 Method for detecting and identifying garbage in elevator based on deep learning
CN112802052A (en) * 2021-01-19 2021-05-14 北京小米移动软件有限公司 Image recognition method and device, electronic equipment and storage medium
CN113055654A (en) * 2021-03-26 2021-06-29 太原师范学院 Method for lossy compression of video stream in edge device
CN115909215A (en) * 2022-12-09 2023-04-04 厦门农芯数字科技有限公司 Edge intrusion early warning method and system based on target detection

Similar Documents

Publication Publication Date Title
CN106372576A (en) Deep learning-based intelligent indoor intrusion detection method and system
CN106803301A (en) A kind of recognition of face guard method and system based on deep learning
Yin et al. Recurrent convolutional network for video-based smoke detection
CN108229362B (en) Binocular face recognition living body detection method based on access control system
CN111832457B (en) Stranger intrusion detection method based on cloud edge cooperation
CN103839346B (en) A kind of intelligent door and window anti-intrusion device and system, intelligent access control system
CN105868735B (en) A kind of preprocess method of track human faces and wisdom health monitor system based on video
CN106919921B (en) Gait recognition method and system combining subspace learning and tensor neural network
CN104504395A (en) Method and system for achieving classification of pedestrians and vehicles based on neural network
WO2019114145A1 (en) Head count detection method and device in surveillance video
CN106339657B (en) Crop straw burning monitoring method based on monitor video, device
CN108416256A (en) The family&#39;s cloud intelligent monitor system and monitoring method of feature based identification
CN105975938A (en) Smart community manager service system with dynamic face identification function
CN103530657B (en) A kind of based on weighting L2 extraction degree of depth study face identification method
CN107947874B (en) Indoor map semantic identification method based on WiFi channel state information
CN210072642U (en) Crowd abnormal behavior detection system based on video monitoring
Zhang et al. A survey on face anti-spoofing algorithms
CN103093526A (en) Intelligent door control system based on scene mode
CN110188715A (en) A kind of video human face biopsy method of multi frame detection ballot
CN110309709A (en) Face identification method, device and computer readable storage medium
CN113139501B (en) Pedestrian multi-attribute identification method combining local area detection and multi-level feature grabbing
CN110956768A (en) Automatic anti-theft device of intelligence house
CN111800617A (en) Intelligent security system based on Internet of things
CN108376237A (en) A kind of house visiting management system and management method based on 3D identifications
CN105652683B (en) Energy control method and its system based on human body infrared characteristic identification human body

Legal Events

Date Code Title Description
C06 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: 20170201