CN103093526A - Intelligent door control system based on scene mode - Google Patents
Intelligent door control system based on scene mode Download PDFInfo
- Publication number
- CN103093526A CN103093526A CN2012105730213A CN201210573021A CN103093526A CN 103093526 A CN103093526 A CN 103093526A CN 2012105730213 A CN2012105730213 A CN 2012105730213A CN 201210573021 A CN201210573021 A CN 201210573021A CN 103093526 A CN103093526 A CN 103093526A
- Authority
- CN
- China
- Prior art keywords
- control module
- image
- module
- control system
- human motion
- 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
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The invention discloses an intelligent door control system based on a scene mode. The intelligent door control system comprises a human body motion image acquisition module, a face image acquisition module, a control unit, a wireless node, a global system for mobile communication (GSM)/general packet radio service (GPRS) module, an electromagnetic valve, an electric curtain and lighting equipment, wherein the control unit is connected with the human body motion image acquisition module, the face image acquisition module and the GSM/GPRS module respectively and is connected with the wireless node through a serial port; and the wireless node is connected with the electromagnetic valve, the electric curtain and the lighting equipment. The door control system based on the scene mode is constructed by an event detection mode and is relatively intelligent, energy-saving and reliable; and application and expansion are convenient.
Description
Technical field
The invention belongs to the technical field of gate control system, particularly, relate to a kind of intelligent access control system based on contextual model.
Background technology
Home intelligence is following development trend, and its objective is to provide convenient, comfortable and safe service according to context aware.Context aware generally is divided into 3 aspects: environment, people's activity and physiological status thereof, realize that the effective detection to above-mentioned aspect is the basis of realizing home intelligence.For traditional gate control system, generally the mode by Proactive authentication enters, as Password Input, and RFID etc., this detachable discriminating means are easily stolen.And but biological attribute has also obtained application due to its identification and security in gate control system, as fingerprint, sound, people's face, iris etc.But the gate control system of the above-mentioned type can not detect for people's activity, therefore can not realize effective context aware, can not satisfy to a certain extent the growth requirement of following home intelligence.
Summary of the invention
Because the defects of prior art, technical matters to be solved by this invention is to provide a kind of intelligent access control system based on contextual model, by the perception to sight, the intelligence of assurance gate control system, energy-conservation, reliable reaching are used and expand conveniently, can adapt to the development of following home intelligence.
For achieving the above object, the invention provides a kind of intelligent access control system based on contextual model, it comprises human motion image acquisition module, man face image acquiring module, control module, radio node, GSM/GPRS module, solenoid valve, electrically driven curtain and light fixture; Wherein, described control module is connected with described human motion image acquisition module, man face image acquiring module and GSM/GPRS module respectively, and is connected with described radio node by serial ports; Described radio node is connected with described solenoid valve, electrically driven curtain and light fixture again;
Wherein, described control module carries out human motion detection and recognition of face to the picture signal of described human motion image acquisition module and the transmission of described man face image acquiring module; Enter the door when target having been detected, within one period time limit, described control module starts the collection to the facial image signal, and the facial image that gathers is carried out feature extraction, then mates with the face database of this locality; After the match is successful, described control module was controlled the unlatching of described solenoid valve, electrically driven curtain and light fixture by described radio node; Go out when target being detected, described control module is controlled described solenoid valve, electrically driven curtain and light fixture by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, described control module sends to owner with the facial image that gathers by described GSM/GPRS module, carries out early warning.
According to the above-mentioned intelligent access control system based on contextual model, wherein: described control module is PC.
According to the above-mentioned intelligent access control system based on contextual model, wherein: described human motion image acquisition module and described man face image acquiring module are camera.
According to the above-mentioned intelligent access control system based on contextual model, wherein: when carrying out human motion and detect, described control module is by setting up the average background model, and human motion image and the described average background model that extracts subtracted each other the human motion prospect that obtains.
Further, according to the above-mentioned intelligent access control system based on contextual model, wherein: when carrying out the human motion detection, for the noise that the extraction human motion image produces, maximum four communicating methods of described control module utilization carry out denoising.
Further, according to the above-mentioned intelligent access control system based on contextual model, wherein: when carrying out the human motion detection, described control module extracts the barycenter that removes the rear prospect of making an uproar by the planar discrete integration, obtains the movement locus of human body; Adopt cosine similarity based method judgement continuous path motion of point direction, realize the motion detection based on barycenter.
According to the above-mentioned intelligent access control system based on contextual model, wherein: when carrying out the detection of people's face, at first described control module constructs the unit orthogonal matrix of local image by svd, obtain the low-dimensional projection vector of local image; Then, described control module carries out projection to the image that gathers, and mates by Euclidean distance with projection vector that the face database of this locality obtains, and the image that the projection vector of described Euclidean distance minimum is corresponding is the facial image of coupling.
Therefore, in the intelligent access control system based on contextual model of the present invention, at first control module carries out foreground extraction, maximum four communicating methods except making an uproar and based on the motion detection of barycenter to the human motion image signal of transmission; Enter the door when target having been detected, within one period time limit, control module starts the collection to the facial image signal; The method that control module utilizes eigenface to the image that gathers carry out feature extraction and with the face database coupling of this locality; After the match is successful, control module is controlled the unlatching of solenoid valve, electrically driven curtain and light fixture by radio node.Leave when target being detected, control module is controlled relevant device by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, the facial image that control module just gathers sends to owner by GPRS, carries out early warning.By the perception to sight, the intelligence of assurance gate control system, energy-conservation and reliable, its application and expansion are convenient, can adapt to well the development of following home intelligence.
Description of drawings
Fig. 1 is the structured flowchart of the intelligent access control system based on contextual model of the present invention;
Fig. 2 is the intelligent access control system workflow diagram based on contextual model of the present invention.
Embodiment
Be described further below with reference to the technique effect of accompanying drawing to design of the present invention, concrete structure and generation, to understand fully purpose of the present invention, feature and effect.
The system of contextual model refers to the intelligence system that can detect and differentiate environment, people's activity and physiological status thereof, and the system of its objective is can carry out different responses by the different event of perception, and the intelligent degree that presents is higher.Detect the detection that triggers people's face by human motion, can reduce on the one hand the calculated amount of system, reach energy-conservation effect, on the other hand, make system can obtain environment event and response thereupon; Facial image has relative uniqueness and stability simultaneously, and belongs to the detection of passive type, is applicable to the application of gate control system.
As shown in Figure 1, the intelligent access control system based on contextual model of the present invention comprises human motion image acquisition module, man face image acquiring module, control module, radio node, solenoid valve, electrically driven curtain and light fixture.Wherein, control module is connected with human motion image acquisition module, man face image acquiring module and GSM/GPRS module respectively, and is connected with radio node by serial ports; Radio node is connected with solenoid valve, electrically driven curtain and light fixture by the bottom radio node again.Particularly, human motion image acquisition module, man face image acquiring are established module for being common camera, and control module is PC.
After camera was fixing, control module first extracted and specifies number human motion image, carries out average and the threshold operation of Pixel-level, and off-line is set up the average background model; After system's operation, control module carries out the background subtracting method with human motion image signal and the background model extracted, obtains the human motion prospect; For extracting the noise that produces, utilize maximum four communicating methods to remove and make an uproar, divide set according to the connection situation of foreground pixel point; After obtaining prospect, extract the barycenter that removes the rear prospect of making an uproar by the planar discrete integration, obtain the movement locus of human body; Adopt cosine similarity based method judgement continuous path motion of point direction, thereby realize the motion detection based on barycenter; Enter when target having been detected, control module starts regularly, and within the time limit of setting, control module starts the collection to the facial image signal; Construct the unit orthogonal matrix of local image by svd, obtain the low-dimensional projection vector of local image data base; Control module carries out projection to the facial image that gathers, and mates by Euclidean distance with projection vector that the face database of this locality obtains; After the match is successful, control module is controlled the unlatching of solenoid valve, electrically driven curtain and light fixture by radio node.Leave when target being detected, control module is controlled solenoid valve, electrically driven curtain and light fixture by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, the facial image that control module just gathers sends to owner by GPRS, carries out early warning.
In the intelligent access control system based on contextual model of the present invention, the specific implementation method that human motion detects is as follows:
At first control module extraction N opens human motion RGB image, first carries out gray processing and histogram equalization and processes, and then carries out pixel average μ (x, y) and threshold tau (x, y) the average background model is set up in computing, wherein p (x, y, i) be the i pictures at the pixel value of the high y coordinate of wide x, μ (x, y) is that the N pictures is at the average pixel value of the high y coordinate of wide x, τ (x, y) is that the N pictures is at the max-thresholds of the high y coordinate of wide x.In the N pictures that gathers, be (x, y) pixel for coordinate, get the N pictures and carry out the pixel value progressive mean, thereby obtain the average image that formed by μ (x, y).
After obtaining the average image, need to obtain the threshold value of correspondence image coordinate, for this reason, more every width image coordinate (x, y) absolute difference of corresponding pixel value and source images, get maximum absolute value differences and be the pixel prospect threshold value that (x, y) puts, this value will be for static foreground extraction.So far, Background Modeling is completed.
After Background Modeling is completed, control module is made poor method with human motion image signal and the background model extracted, obtain preliminary human motion prospect: the threshold value that surpasses respective pixel point when the corresponding absolute difference of background model of the human motion image signal that extracts and foundation, the pixel value of foreground image corresponding coordinate is set as 255, be white point, otherwise be stain.
After obtaining initial prospect, first it is carried out the 15*15 mean filter, be about to image space and be divided into a plurality of 15*15 grid, each grid pixel value is set to whole grid pixel average, then setting threshold, pixel value to each grid carries out binary conversion treatment, with the isolated stain in the inside of eliminating prospect.After early stage, the image pre-service was completed, for extracting the noise that produces, utilize following maximum four communicating methods to remove and make an uproar.
An initial newly-built set, according to image from push up downwards, from left to right traversal is extracted the initial motion prospect, if the pixel of traversal is communicated with (pixel value is 255) with the pixel in its top or left side, it is joined corresponding set, otherwise a newly-built set continues traversal; After once traversal is completed, continue the secondary traversal, if p is (x, y) with p (x-1, y+1) pixel value is not 0 and corresponding affiliated set difference (wherein p (x, y) presentation video two values matrix is at the pixel value of the high y coordinate of wide x), with p (x-1, y+1) under, the pixel of set adds in the affiliated set of p (x, y); Set is respectively gathered the pixel number after integrating and completing, and obtains the set over assign thresholds pixel number maximum, and the pixel value of this set-inclusion pixel coordinate is put 255, obtains except the prospect after making an uproar.
Obtain extracting except after the prospect after making an uproar the barycenter that removes the prospect after making an uproar by the planar discrete integration, and then obtain the movement locus of human body, as follows, m
10, m
01And m
00Square and whole " quality " of foreground image on level, vertical direction that represents respectively extraction.Calculation expression is as follows, and wherein x, y are respectively the wide and high of image.Square on level, vertical direction is
Whole " quality "
Obtain the square of image level and vertical direction by following formula after, the image level square is just obtained the human motion barycenter divided by the image " quality " of integral body respectively with vertical square, barycenter is namely arranged
After extracting mass center of human body, the sampling center of mass point was continued at the interval in one second, thereby obtained movement locus, and rear employing cosine similarity based method judgement continuous path motion of point direction realizes the motion detection based on barycenter.The vectorial A and the cartesian coordinate system all directions vector of unit length E that obtain according to adjacent center of mass point
1=(1,0), E
2=(0,1), E
3=(1,0), E
4The angle theta of=(0 ,-1), wherein E
1, E
2, E
3, E
4For take initial point as starting point, point to respectively positive x axle, positive y axle, negative x axle, the vector of unit length of negative y axle.If θ is (E
1, A) ∈ [0,20 °], judge that human body moves right; If θ is (E
2, A) ∈ [0,45 °], judge that human body moves upward; If θ is (E
3, A) ∈ [0,20 °] judges that human body is to left movement; If θ is (E
4, A) ∈ [0,45 °], judge that human body moves downward; If other values, direction can not detect.
After detecting perfect person's running body direction, control module starts the collection to the facial image signal, carry out recognition of face, and the specific implementation method of recognition of face is described below:
At first construct the unit orthogonal matrix of local image by svd, have before this following theorem to exist.
Known to matrix A
Row * MIf its order equals r, wherein row=m*n, wherein m and n represent respectively to obtain width and the height of image, and M represents the number of local image library, exists: the order that matrix A multiply by its transposition equals the order of matrix A, be equal to r, by with co-relation, be convenient to matrix A is carried out svd.
By the svd of matrix A, A=[φ is arranged
1, φ
2..., φ
M]=U Λ V
T
Wherein
And Γ
iBe the m*n dimensional vector that i width image launches by row,
The average image for M image; U and V are respectively the orthogonal characteristic vector of matrix A and A transposition product or A transposition and A product, and U=[u
0, u
1, u
2..., u
M*n-1], V=[v
0, v
1, v
2..., v
M-1], wherein Λ is the diagonal matrix take r singular value of its matrix A as diagonal entry, specifically is expressed as
In Λ, the singular value of matrix A
Along with i increases and successively decreases, it is the singular value of matrix A, and A is arranged
TAv
i=λ
iv
i, AA
Tu
i=λ
i(Av
i), by this method of indirectly asking, greatly reduce the calculated amount of finding the solution of U.
After obtaining the proper vector of step, can carry out further dimension-reduction treatment with further reduction complexity computing time to the orthogonal characteristic vector U of unit, method is to get larger eigenvalue λ
iCharacteristic of correspondence vector like this can be when projection pulls open the distance of the projection vector of each image, greatly reduces on the one hand the computation complexity (projection vector is reduced to the k+1 dimension from the m*n dimension) of images match; Very little for coupling and the classification performance impact of image on the other hand.The proper vector number (k+1) of choosing is weighed by selected characteristic value element with in the ratio of whole eigenwert summation, satisfy make the eigenwert element and with the minimum value of whole eigenwert summation greater than certain constant.
Wherein constant generally gets 0.99, the facial image matching complete of expression 99%.
And formula (1) has just illustrated at U
kOn process and the principle of projection P dimensionality reduction, projection vector is reduced to the k+1 dimension from the m*n dimension.
After obtaining the low-dimensional projection vector of local image data base, know simultaneously p
iAnd
As follows:
Thereby learn that eigenwert is larger, corresponding each dimension distance of each image projection vector is also larger, is beneficial to the classification of projection vector, namely is convenient to the last identification of facial image.
After control module carried out projection to the image that gathers, the projection vector that obtains with the face database of this locality mated in image 1～M by Euclidean distance, and image corresponding to projection vector of getting the Euclidean distance minimum is the facial image of coupling.
The step of front has realized that respectively human motion detects and people's face of order detects, and the below is exactly the whole experimentation on this two parts bases.As shown in Figure 2: during work, control module first extracts and specifies number human motion image, carries out average and the threshold operation of Pixel-level, and off-line is set up the average background model; After system's operation, control module carries out the background subtraction method with human motion image signal and the background model extracted, obtains the human motion prospect; For extracting the noise that produces, utilize maximum four communicating methods to remove and make an uproar, divide set according to the connection situation of foreground pixel point; Obtain extracting except after the prospect after making an uproar the barycenter that removes the rear prospect of making an uproar by the planar discrete integration, obtain the movement locus of human body; Adopt cosine similarity based method judgement continuous path motion of point direction, thereby realize the motion detection based on barycenter; Enter when target having been detected, control module starts regularly (6s), and in the limit, control module starts the collection to the facial image signal at this moment; By constructing the unit orthogonal matrix of local image, obtain the low-dimensional projection vector of local image data base; Control module carries out projection to the image that gathers, and mates by Euclidean distance with projection vector that the face database of this locality obtains; After the match is successful, control module is controlled the unlatching of solenoid valve, electrically driven curtain and light fixture by radio node, and after 4s, system circulates again.Leave when target being detected, control module is controlled relevant device by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, control module sends to owner with matching image by GPRS, carries out early warning;
The objective of the invention is to build a kind of intelligent access control system based on contextual model.Utilize background subtraction method, maximum four connected region methods and the realization of barycenter extraction method to extraction, denoising and the motion detection in human motion zone; Utilize the eigenface method to carry out feature extraction and identification to facial image; Thereby realize the perception to sight.After the event detection success, control module is controlled the unlatching of solenoid valve, electrically driven curtain and light fixture by radio node.Leave when target being detected, control module is controlled relevant device by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, control module sends to owner with matching image by GPRS, carries out early warning.By the perception to above-mentioned sight, guarantee the intelligence, energy-conservation and reliable of system, its application and expansion are convenient, can adapt to well development and the demand of following Smart Home.
More than describe preferred embodiment of the present invention in detail.The ordinary skill that should be appreciated that this area need not creative work and just can design according to the present invention make many modifications and variations.Therefore, all technician in the art all should be in the determined protection domain by claims under this invention's idea on the basis of existing technology by the available technical scheme of logical analysis, reasoning, or a limited experiment.
Claims (7)
1. the intelligent access control system based on contextual model, is characterized in that, comprises human motion image acquisition module, man face image acquiring module, control module, radio node, GSM/GPRS module, solenoid valve, electrically driven curtain and light fixture; Wherein, described control module is connected with described human motion image acquisition module, man face image acquiring module and GSM/GPRS module respectively, and is connected with described radio node by serial ports; Described radio node is connected with described solenoid valve, electrically driven curtain and light fixture again;
Wherein, described control module carries out human motion detection and recognition of face to the picture signal of described human motion image acquisition module and the transmission of described man face image acquiring module; Enter the door when target having been detected, within one period time limit, described control module starts the collection to the facial image signal, and the facial image that gathers is carried out feature extraction, then mates with the face database of this locality; After the match is successful, described control module was controlled the unlatching of described solenoid valve, electrically driven curtain and light fixture by described radio node; Go out when target being detected, described control module is controlled described solenoid valve, electrically driven curtain and light fixture by radio node and is closed; Surpass threshold value if mate unsuccessful number of times, described control module sends to owner with the facial image that gathers by described GSM/GPRS module, carries out early warning.
2. the intelligent access control system based on contextual model according to claim 1, it is characterized in that: described control module is PC.
3. the intelligent access control system based on contextual model according to claim 1, it is characterized in that: described human motion image acquisition module and described man face image acquiring module are camera.
4. the intelligent access control system based on contextual model according to claim 1, it is characterized in that: when carrying out the human motion detection, described control module is by setting up the average background model, and human motion image and the described average background model that extracts subtracted each other the human motion prospect that obtains.
5. the intelligent access control system based on contextual model according to claim 4 is characterized in that: when carrying out human motion and detect, for the noise that extracts human motion image and produce, maximum four communicating methods of described control module utilization carry out denoising.
6. the intelligent access control system based on contextual model according to claim 5 is characterized in that: when carrying out human motion and detect, described control module extracts barycenter except the rear prospect of making an uproar by the planar discrete integration, obtains the movement locus of human body; Adopt cosine similarity based method judgement continuous path motion of point direction, realize the motion detection based on barycenter.
7. the intelligent access control system based on contextual model according to claim 1, it is characterized in that: when carrying out the detection of people's face, at first described control module constructs the unit orthogonal matrix of local image by svd, obtain the low-dimensional projection vector of local image; Then, described control module carries out projection to the image that gathers, and mates by Euclidean distance with projection vector that the face database of this locality obtains, and the image that the projection vector of described Euclidean distance minimum is corresponding is the facial image of coupling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012105730213A CN103093526A (en) | 2012-12-25 | 2012-12-25 | Intelligent door control system based on scene mode |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012105730213A CN103093526A (en) | 2012-12-25 | 2012-12-25 | Intelligent door control system based on scene mode |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103093526A true CN103093526A (en) | 2013-05-08 |
Family
ID=48206055
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012105730213A Pending CN103093526A (en) | 2012-12-25 | 2012-12-25 | Intelligent door control system based on scene mode |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103093526A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103258191A (en) * | 2013-05-15 | 2013-08-21 | 苏州福丰科技有限公司 | Community access control system based on face recognition |
| CN104093007A (en) * | 2014-08-01 | 2014-10-08 | 北京奇虎科技有限公司 | Mobile communication equipment, video monitoring system and video monitoring method |
| CN104880953A (en) * | 2015-05-29 | 2015-09-02 | 四川长虹电器股份有限公司 | Control method and intelligent household control device |
| CN105334765A (en) * | 2014-08-06 | 2016-02-17 | 北大方正集团有限公司 | Power circuit control method and device |
| CN105804436A (en) * | 2016-03-21 | 2016-07-27 | 青岛源之林农业科技开发有限公司 | Toilet dry and wet region isolation system with drying function and toilet dry and wet region isolation method |
| CN106411670A (en) * | 2016-10-11 | 2017-02-15 | 深圳万发创新进出口贸易有限公司 | Smart home system |
| CN108257276A (en) * | 2018-02-26 | 2018-07-06 | 宁夏煜隆科技有限公司 | Entrance guard controlling method and system |
| CN108307120A (en) * | 2018-05-11 | 2018-07-20 | 优视科技有限公司 | Image capturing method, device and electric terminal |
| CN109656161A (en) * | 2018-11-12 | 2019-04-19 | 厦门快商通信息技术有限公司 | A kind of smart meeting room control system and its control method |
| CN110458062A (en) * | 2019-07-30 | 2019-11-15 | 深圳市商汤科技有限公司 | Face identification method and device, electronic equipment and storage medium |
| CN112309022A (en) * | 2020-10-22 | 2021-02-02 | 天津时利和企业管理咨询有限公司 | Multifunctional docking station of face recognition terminal |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289220A (en) * | 2010-10-21 | 2011-12-21 | 青岛科技大学 | Intelligent household perception and control system |
| CN102647318A (en) * | 2012-05-02 | 2012-08-22 | 江苏下一代广电网物联网研究中心有限公司 | Intelligent household system controlled through human characteristic information and working method thereof |
| CN102830675A (en) * | 2012-08-28 | 2012-12-19 | 北京理工大学 | Intelligent home robot system based on GIS (geographic information system) |
-
2012
- 2012-12-25 CN CN2012105730213A patent/CN103093526A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289220A (en) * | 2010-10-21 | 2011-12-21 | 青岛科技大学 | Intelligent household perception and control system |
| CN102647318A (en) * | 2012-05-02 | 2012-08-22 | 江苏下一代广电网物联网研究中心有限公司 | Intelligent household system controlled through human characteristic information and working method thereof |
| CN102830675A (en) * | 2012-08-28 | 2012-12-19 | 北京理工大学 | Intelligent home robot system based on GIS (geographic information system) |
Non-Patent Citations (1)
| Title |
|---|
| LONG XIAO,ET AL.: "A Context-aware Entrance Guard in Smart Home: an Event-driven Application Based on the Human Motion and Face Recognition", 《PROCEEDINGS OF THE 5TH INTERNATIONAL CONFERENCE ON AUTOMATION,ROBOTICS AND APPLICATIONS》, 8 December 2011 (2011-12-08) * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103258191A (en) * | 2013-05-15 | 2013-08-21 | 苏州福丰科技有限公司 | Community access control system based on face recognition |
| CN104093007A (en) * | 2014-08-01 | 2014-10-08 | 北京奇虎科技有限公司 | Mobile communication equipment, video monitoring system and video monitoring method |
| CN104093007B (en) * | 2014-08-01 | 2018-05-29 | 北京奇虎科技有限公司 | Mobile communication equipment, video monitoring system and video frequency monitoring method |
| CN105334765A (en) * | 2014-08-06 | 2016-02-17 | 北大方正集团有限公司 | Power circuit control method and device |
| CN104880953A (en) * | 2015-05-29 | 2015-09-02 | 四川长虹电器股份有限公司 | Control method and intelligent household control device |
| CN105804436A (en) * | 2016-03-21 | 2016-07-27 | 青岛源之林农业科技开发有限公司 | Toilet dry and wet region isolation system with drying function and toilet dry and wet region isolation method |
| CN106411670A (en) * | 2016-10-11 | 2017-02-15 | 深圳万发创新进出口贸易有限公司 | Smart home system |
| CN108257276A (en) * | 2018-02-26 | 2018-07-06 | 宁夏煜隆科技有限公司 | Entrance guard controlling method and system |
| CN108307120A (en) * | 2018-05-11 | 2018-07-20 | 优视科技有限公司 | Image capturing method, device and electric terminal |
| CN108307120B (en) * | 2018-05-11 | 2020-07-17 | 阿里巴巴（中国）有限公司 | Image shooting method and device and electronic terminal |
| CN109656161A (en) * | 2018-11-12 | 2019-04-19 | 厦门快商通信息技术有限公司 | A kind of smart meeting room control system and its control method |
| CN110458062A (en) * | 2019-07-30 | 2019-11-15 | 深圳市商汤科技有限公司 | Face identification method and device, electronic equipment and storage medium |
| CN112309022A (en) * | 2020-10-22 | 2021-02-02 | 天津时利和企业管理咨询有限公司 | Multifunctional docking station of face recognition terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103093526A (en) | Intelligent door control system based on scene mode | |
| Kliper-Gross et al. | Motion interchange patterns for action recognition in unconstrained videos | |
| Charfi et al. | Optimized spatio-temporal descriptors for real-time fall detection: comparison of support vector machine and Adaboost-based classification | |
| CN106803301A (en) | A kind of recognition of face guard method and system based on deep learning | |
| CN106372576A (en) | Deep learning-based intelligent indoor intrusion detection method and system | |
| CN101226597B (en) | Method and system for recognizing nights pedestrian based on thermal infrared gait | |
| CN102903119B (en) | A kind of method for tracking target and device | |
| CN103198303B (en) | A kind of gender identification method based on facial image | |
| CN109145742A (en) | A kind of pedestrian recognition method and system | |
| CN103530657B (en) | A kind of based on weighting L2 extraction degree of depth study face identification method | |
| CN105701467A (en) | Many-people abnormal behavior identification method based on human body shape characteristic | |
| US20180173940A1 (en) | System and method for matching an object in captured images | |
| CN102609724B (en) | Method for prompting ambient environment information by using two cameras | |
| CN103268497A (en) | Gesture detecting method for human face and application of gesture detecting method in human face identification | |
| Soni et al. | Hybrid meta-heuristic algorithm based deep neural network for face recognition | |
| CN103971384B (en) | Node cooperation target tracking method of wireless video sensor | |
| Ta et al. | Recognizing and localizing individual activities through graph matching | |
| CN105354856A (en) | Human matching and positioning method and system based on MSER and ORB | |
| Charfi et al. | Optimised spatio-temporal descriptors for real-time fall detection: comparison of SVM and Adaboost based classification | |
| CN102779270A (en) | Target clothing image extraction method aiming at shopping image search | |
| Khan et al. | Multiple human detection in depth images | |
| CN110516533A (en) | A kind of pedestrian based on depth measure discrimination method again | |
| CN105224918A (en) | Based on the gait recognition method of bilinearity joint sparse discriminatory analysis | |
| CN109815780A (en) | A kind of high-precision fingerprint identification method and system based on image procossing | |
| CN103714340B (en) | Self-adaptation feature extracting method based on image partitioning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130508 |