CN108710857A - People's vehicle recognition methods based on infrared light filling and device - Google Patents
People's vehicle recognition methods based on infrared light filling and device Download PDFInfo
- Publication number
- CN108710857A CN108710857A CN201810495514.7A CN201810495514A CN108710857A CN 108710857 A CN108710857 A CN 108710857A CN 201810495514 A CN201810495514 A CN 201810495514A CN 108710857 A CN108710857 A CN 108710857A
- Authority
- CN
- China
- Prior art keywords
- face
- vehicle
- region
- recognition
- infrared light
- 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.)
- Granted
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
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Abstract
The present invention provides people's vehicle recognition methods based on infrared light filling, this method includes：Acquire scene image；The luminance mean value of scene image is calculated, if luminance mean value is less than first threshold, opens infrared light filling, and resurvey scene image；Vehicle detection and car plate detection are carried out to scene image, obtain vehicle region and license plate area；If there is license plate area, Car license recognition is carried out to license plate area, obtains license plate recognition result；If there is no license plate area, then vehicle cab recognition is carried out to vehicle region, obtains vehicle cab recognition result；Vehicle window region is set, Face datection is carried out to vehicle window region, human face region is obtained, human face region is identified, obtains face recognition result；Export car plate or vehicle cab recognition result and face recognition result.Compared with existing people's vehicle identification technology, discrimination can be improved by infrared light filling, and on human eye without influence.
Description
Technical field
The present invention relates to the parkings of image procossing, video monitoring and wisdom, the more particularly to recognition methods of people's vehicle and device.
Background technology
With the development of economic technology, motor vehicles it is growing day by day, traditional labor management mode can not increasingly expire
The actual needs of foot.At the same time, with the development of science and technology, Intelligent traffic management systems is just increasingly mature, substitution is traditional gradually
Labor management mode.License plate recognition technology is the basis for realizing intelligent traffic administration system, is increasingly focused on.
In recent years, in order to improve the safety in parking lot, more and more parking lot managements will ask for help vehicle identification integration.
However in actual scene, there are night illumination is relatively low, the identification of face and vehicle is influenced.People's vehicle identifying schemes usually exist at present
Night uses gas flashing light light filling, and gas flashing light light filling can not only influence driver's human eye, also will produce light pollution
Problem.
In conclusion at present there is an urgent need to propose that a kind of solution low-light (level) influences, and human eye is not influenced, it not will produce light
People's vehicle recognition methods of pollution.
Invention content
In view of this, it is a primary object of the present invention to realize that people's vehicle identifies, and discrimination is high.
In order to achieve the above objectives, the first aspect according to the invention provides people's vehicle identification side based on infrared light filling
Method, this method include：
First step acquires scene image；
Second step calculates the luminance mean value of scene image, if luminance mean value is less than first threshold, opens infrared benefit
Light, and resurvey scene image；
Third step carries out vehicle detection and car plate detection to scene image, obtains vehicle region and license plate area；
Four steps carries out Car license recognition to license plate area, obtains license plate recognition result if there is license plate area；Such as
License plate area is not present in fruit, then carries out vehicle cab recognition to vehicle region, obtains vehicle cab recognition result；
5th step, setting vehicle window region carry out Face datection to vehicle window region, human face region are obtained, to human face region
It is identified, obtains face recognition result；
6th step exports car plate or vehicle cab recognition result and face recognition result.
Further, the 5th step includes：
Vehicle window zone enactment steps, according to the position of license plate area, setting vehicle window region；If there is no license plate area,
According to the position of vehicle region, setting vehicle window region；
Face datection step obtains human face region using method for detecting human face from vehicle window region；
Recognition of face step carries out the human face data logged in human face region and database using face identification method
It compares, if compared unanimously, using the user information in database as the face information of human face region, differs if compared
It causes, is then recorded as visitor, work out visitor in the database and number and store visitor's face information.
Further, the vehicle window zone enactment steps include：
Vehicle window setting procedure based on car plate obtains left margin x=pl, the right side of license plate area if there is license plate area
Boundary x=pr, coboundary y=pt, lower boundary y=pb, set the left margin in vehicle window region asRight margin isCoboundary isLower boundary isWpFor car plate area
The width in domain, W are the width of scene image, 3 < λ 2 of λ；
Vehicle window setting procedure based on vehicle, if there is no license plate area, the upper subregion for obtaining vehicle region is made
For vehicle window region.
Further, the recognition of face step may include：Marked face sample image is chosen, to convolutional Neural
Network is trained, and obtains trained face Network Recognition model；Using face Network Recognition model to human face region and number
It is compared according to the user's human face data logged in library, if compared unanimously, using the user information in database as face
The face information in region is recorded as visitor if comparison is inconsistent, works out visitor in the database and numbers and store visitor people
Face information.
The marked face sample image includes：Different angle, different brightness, different postures under non-infrared light compensating lamp
The facial image for having marked classification, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
Further, the recognition of face step can also include：
First network identification model training step chooses marked Generic face sample image, to the first convolutional Neural
Network is trained, and obtains trained first network identification model；
Second Network Recognition model classification step chooses marked infrared face sample image, to the second convolutional Neural
Network is trained, and obtains trained second Network Recognition model；
Human face region identification step, when not opening infrared light filling, using first network identification model to human face region with
The user's human face data logged in database is compared, when opening infrared light filling, using the second Network Recognition model pair with
The user's human face data logged in database is compared；If compared unanimously, using the user information in database as people
The face information in face region is recorded as visitor if comparison is inconsistent, works out visitor in the database and numbers and store visitor
Face information.
The marked Generic face sample image includes：Different angle, different brightness, difference under non-infrared light compensating lamp
The facial image for having marked classification of posture.The marked infrared face sample image includes：It is different under infrared light compensating lamp
The facial image for having marked classification of angle, different postures.
Other side according to the invention, provides people's vehicle identification device based on infrared light filling, which includes：
Scene image acquisition module, for acquiring scene image；
Infrared light filling control module, the luminance mean value for calculating scene image, if luminance mean value is less than first threshold,
Infrared light filling is then opened, and resurveys scene image；
Vehicle license plate detection module, for carrying out vehicle detection and car plate detection to scene image, obtain vehicle region and
License plate area；
Car plate vehicle cab recognition module, for if there is license plate area, carrying out Car license recognition to license plate area, obtaining car plate
Recognition result；If there is no license plate area, then vehicle cab recognition is carried out to vehicle region, obtains vehicle cab recognition result；
Vehicle window region face recognition module carries out Face datection to vehicle window region, obtains face for setting vehicle window region
Human face region is identified in region, obtains face recognition result；
People's vehicle recognition result output module, for exporting car plate or vehicle cab recognition result and face recognition result.
Further, vehicle window region face recognition module includes：
Vehicle window region setting module, for the position according to license plate area, setting vehicle window region；If there is no car plate area
Domain, according to the position of vehicle region, setting vehicle window region；
Face detection module obtains human face region for using method for detecting human face from vehicle window region；
Face recognition module, for using face identification method, to the human face data logged in human face region and database
It is compared, if compared unanimously, using the user information in database as the face information of human face region, if compared not
Unanimously, then it is recorded as visitor, visitor is worked out in the database and numbers and store visitor's face information.
Further, vehicle window region setting module includes：
Vehicle window setting module based on car plate, for if there is license plate area, obtaining the left margin x=of license plate area
Pl, right margin x=pr, coboundary y=pt, lower boundary y=pb, set the left margin in vehicle window region asRight margin isCoboundary isLower boundary isWpFor car plate area
The width in domain, W are the width of scene image, 3 < λ 2 of λ；
Vehicle window setting module based on vehicle, for if there is no license plate area, obtaining the top subregion of vehicle region
Domain is as vehicle window region.
Further, the face recognition module may include：For choosing marked face sample image, to convolution
Neural network is trained, and obtains trained face Network Recognition model；For using face Network Recognition model to face
Region is compared with the user's human face data logged in database, if compared unanimously, by the user information in database
As the face information of human face region, if comparison is inconsistent, it is recorded as visitor, visitor is worked out in the database and numbers and deposit
Store up visitor's face information.
The marked face sample image includes：Different angle, different brightness, different postures under non-infrared light compensating lamp
The facial image for having marked classification, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
Further, the face recognition module can also include：
First network identification model training module, for choosing marked Generic face sample image, to the first convolution
Neural network is trained, and obtains trained first network identification model；
Second Network Recognition category of model module, for choosing marked infrared face sample image, to the second convolution
Neural network is trained, and obtains trained second Network Recognition model；
Human face region identification module, for when not opening infrared light filling, using first network identification model to face area
Domain is compared with the user's human face data logged in database, when opening infrared light filling, using the second Network Recognition model
It pair is compared with the user's human face data logged in database；If consistent for comparing, the user in database is believed
It ceases the face information as human face region and is recorded as visitor if comparison is inconsistent, work out visitor's number in the database simultaneously
Store visitor's face information.
The marked Generic face sample image includes：Different angle, different brightness, difference under non-infrared light compensating lamp
The facial image for having marked classification of posture.The marked infrared face sample image includes：It is different under infrared light compensating lamp
The facial image for having marked classification of angle, different postures.
Compared with existing people's vehicle identification technology, the recognition methods of people's vehicle and device of the invention based on infrared light filling according to
Picture quality carries out infrared light filling, the accuracy rate of night people's vehicle identification can be improved, simultaneously because using infrared light filling, to human eye
Without influence, light pollution will not be led to the problem of.
Description of the drawings
Fig. 1 shows the flow chart of people's vehicle recognition methods according to the invention based on infrared light filling.
The frame diagram of Fig. 2 shows according to the invention people's vehicle identification device based on infrared light filling.
Specific implementation mode
To enable those skilled in the art to further appreciate that structure, feature and the other purposes of the present invention, in conjunction with institute
Detailed description are as follows for attached preferred embodiment, and illustrated preferred embodiment is only used to illustrate the technical scheme of the present invention, and is not limited
The fixed present invention.
Fig. 1 gives the flow chart of people's vehicle recognition methods according to the invention based on infrared light filling.As shown in Figure 1, pressing
Include according to people's vehicle recognition methods based on infrared light filling of the invention：
First step S1 acquires scene image；
Second step S2 calculates the luminance mean value of scene image, if luminance mean value is less than first threshold, opens infrared
Light filling, and resurvey scene image；
Third step S3 carries out vehicle detection and car plate detection to scene image, obtains vehicle region and license plate area；
Four steps S4 carries out Car license recognition to license plate area, obtains license plate recognition result if there is license plate area；
If there is no license plate area, then vehicle cab recognition is carried out to vehicle region, obtains vehicle cab recognition result；
5th step S5, setting vehicle window region carry out Face datection to vehicle window region, human face region are obtained, to face area
Domain is identified, and obtains face recognition result；
6th step S6 exports car plate or vehicle cab recognition result and face recognition result.
The first step S1 can acquire scene image by existing image capture device or device.Embodiment,
Gun-type camera is installed at the entrance in parking lot, for acquiring the scene image at entrance.
Further, the value range of the first threshold is 90~110.The infrared light filling passes through existing infrared benefit
Light lamp is realized.Embodiment, the second step S2 are：The average value of the brightness value of all pixels point in scene image is calculated, such as
The average value of the brightness value of fruit all pixels point is less than 100, then controls infrared light compensating lamp unlatching, while resurveying scene graph
Picture.
Further, the third step S3 includes：Vehicle detection step S31, using vehicle checking method, to scene graph
As carrying out vehicle detection, vehicle detection region is obtained；Car plate detection step S32, using detection method of license plate, to scene image into
Row car plate detection obtains car plate detection region.The vehicle checking method and detection method of license plate can pass through existing vehicle
Detection technique and car plate detection technology are realized.
The four steps S4 may be used existing license plate recognition technology and realize Car license recognition, be known using existing vehicle
Other technology realizes vehicle cab recognition.
Further, the 5th step S5 includes：
Vehicle window zone enactment steps S51, according to the position of license plate area, setting vehicle window region；If there is no car plate area
Domain, according to the position of vehicle region, setting vehicle window region；
Face datection step S52 obtains human face region using method for detecting human face from vehicle window region；
Recognition of face step S53, using face identification method, to the human face data that is logged in human face region and database into
Row compares, if compared unanimously, using the user information in database as the face information of human face region, differs if compared
It causes, is then recorded as visitor, work out visitor in the database and number and store visitor's face information.
Further, the vehicle window zone enactment steps S51 includes：
Vehicle window setting procedure S511 based on car plate obtains the left margin x=of license plate area if there is license plate area
Pl, right margin x=pr, coboundary y=pt, lower boundary y=pb, set the left margin in vehicle window region asRight margin isCoboundary
ForLower boundary isWpFor car plate area
The width in domain, W are the width of scene image, 3 < λ 2 of λ；
Vehicle window setting procedure S512 based on vehicle obtains the top subregion of vehicle region if there is no license plate area
Domain is as vehicle window region.
Further, 1 ∈ [ of the λ;1.2,1.8], 2 ∈ [ of λ;4.2,4.8], 3 ∈ [ of λ;0.3,0.8].Embodiment, λ 1 are selected as
1.5, λ 2, which are selected as 4.5, λ 3, is selected as 0.5.
Embodiment, the vehicle window setting procedure S512 based on vehicle are：If there is no license plate area, vehicle area is obtained
Coboundary y=vt, the lower boundary y=vb in domain, the left margin in setting vehicle window region, right margin, coboundary are respectively vehicle region
Left margin, right margin, coboundary, set the lower boundary in vehicle window region as
Face identification method can be existing face identification method or face described in the recognition of face step S53
Comparison method.Further, the face identification method uses the face identification method based on convolutional neural networks.
Further, the recognition of face step S53 may include：Marked face sample image is chosen, to convolution
Neural network is trained, and obtains trained face Network Recognition model；Using face Network Recognition model to human face region
Be compared with the user's human face data logged in database, if compare it is consistent, using the user information in database as
The face information of human face region is recorded as visitor if comparison is inconsistent, works out visitor in the database and numbers and store visit
Objective face information.
The marked face sample image includes：Different angle, different brightness, different postures under non-infrared light compensating lamp
The facial image for having marked classification, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
Further, the recognition of face step S53 can also include：
First network identification model training step S531, chooses marked Generic face sample image, to the first convolution
Neural network is trained, and obtains trained first network identification model；
Second Network Recognition model classification step S532, chooses marked infrared face sample image, to the second convolution
Neural network is trained, and obtains trained second Network Recognition model；
Human face region identification step S533, when not opening infrared light filling, using first network identification model to face area
Domain is compared with the user's human face data logged in database, when opening infrared light filling, using the second Network Recognition model
It pair is compared with the user's human face data logged in database；If compared unanimously, the user information in database is made
It is recorded as visitor if comparison is inconsistent for the face information of human face region, visitor is worked out in the database and numbers and store
Visitor's face information.
The marked Generic face sample image includes：Different angle, different brightness, difference under non-infrared light compensating lamp
The facial image for having marked classification of posture.The marked infrared face sample image includes：It is different under infrared light compensating lamp
The facial image for having marked classification of angle, different postures.
The convolutional neural networks, the first convolutional neural networks and second convolutional neural networks can be existing volume
Product neural network.
Further, the 6th step S6 includes：If there is license plate recognition result, then export license plate recognition result and
Face recognition result；If there is no license plate recognition result, then vehicle cab recognition result and face recognition result are exported.
Fig. 2 gives the frame diagram of people's vehicle identification device according to the invention based on infrared light filling.As shown in Fig. 2, pressing
Include according to people's vehicle identification device based on infrared light filling of the invention：
Scene image acquisition module 1, for acquiring scene image；
Infrared light filling control module 2, the luminance mean value for calculating scene image, if luminance mean value is less than the first threshold
Value, then open infrared light filling, and resurvey scene image；
Vehicle license plate detection module 3, for carrying out vehicle detection and car plate detection to scene image, obtain vehicle region and
License plate area；
Car plate vehicle cab recognition module 4, for if there is license plate area, carrying out Car license recognition to license plate area, obtaining vehicle
Board recognition result；If there is no license plate area, then vehicle cab recognition is carried out to vehicle region, obtains vehicle cab recognition result；
Vehicle window region face recognition module 5 carries out Face datection to vehicle window region, obtains people for setting vehicle window region
Face region, is identified human face region, obtains face recognition result；
People's vehicle recognition result output module 6, for exporting car plate or vehicle cab recognition result and face recognition result.
Further, the value range of the first threshold is 90~110.The scene image acquisition module 1 is existing
Image capture device or device.
Further, the vehicle license plate detection module 3 includes：Vehicle detection module 31, for using vehicle detection side
Method carries out vehicle detection to scene image, obtains vehicle detection region；Car plate detection module 32, for using car plate detection side
Method carries out car plate detection to scene image, obtains car plate detection region.
Existing car license recognition equipment may be used in the car plate vehicle cab recognition module 4 or device realizes Car license recognition,
Vehicle cab recognition is realized using existing vehicle cab recognition equipment or device.
Further, vehicle window region face recognition module 5 includes：
Vehicle window region setting module 51, for the position according to license plate area, setting vehicle window region；If there is no car plate
Region, according to the position of vehicle region, setting vehicle window region；
Face detection module 52 obtains human face region for using method for detecting human face from vehicle window region；
Face recognition module 53, for using face identification method, to the face number logged in human face region and database
According to being compared, if compared unanimously, using the user information in database as the face information of human face region, if compared
It is inconsistent, then it is recorded as visitor, visitor is worked out in the database and numbers and store visitor's face information.
Further, vehicle window region setting module 51 includes：
Vehicle window setting module 511 based on car plate, for if there is license plate area, obtaining the left margin x of license plate area
=pl, right margin x=pr, coboundary y=pt, lower boundary y=pb, set the left margin in vehicle window region asRight margin isCoboundary
ForLower boundary isWpFor car plate area
The width in domain, W are the width of scene image, 3 < λ 2 of λ；
Vehicle window setting module 512 based on vehicle, for if there is no license plate area, obtaining the upper part of vehicle region
Region is as vehicle window region.
Further, 1 ∈ [ of the λ;1.2,1.8], 2 ∈ [ of λ;4.2,4.8], 3 ∈ [ of λ;0.3,0.8].
Further, the face recognition module 53 may include：For choosing marked face sample image, to volume
Product neural network is trained, and obtains trained face Network Recognition model；For using face Network Recognition model to people
Face region is compared with the user's human face data logged in database, if compared unanimously, the user in database is believed
It ceases the face information as human face region and is recorded as visitor if comparison is inconsistent, work out visitor's number in the database simultaneously
Store visitor's face information.
The marked face sample image includes：Different angle, different brightness, different postures under non-infrared light compensating lamp
The facial image for having marked classification, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
Further, the face recognition module 53 can also include：
First network identification model training module 531, for choosing marked Generic face sample image, to the first volume
Product neural network is trained, and obtains trained first network identification model；
Second Network Recognition category of model module 532, for choosing marked infrared face sample image, to volume Two
Product neural network is trained, and obtains trained second Network Recognition model；
Human face region identification module 533, for when not opening infrared light filling, using first network identification model to face
Region is compared with the user's human face data logged in database, when opening infrared light filling, using the second Network Recognition mould
Type pair is compared with the user's human face data logged in database；If consistent for comparing, by the user in database
Face information of the information as human face region is recorded as visitor if comparison is inconsistent, works out visitor's number in the database
And store visitor's face information.
The marked Generic face sample image includes：Different angle, different brightness, difference under non-infrared light compensating lamp
The facial image for having marked classification of posture.The marked infrared face sample image includes：It is different under infrared light compensating lamp
The facial image for having marked classification of angle, different postures.
Further, people's vehicle recognition result output module 6 includes：It is for if there is license plate recognition result, then defeated
Go out license plate recognition result and face recognition result；If there is no license plate recognition result, then vehicle cab recognition result and face are exported
Recognition result.
Compared with existing people's vehicle identification technology, the recognition methods of people's vehicle and device of the invention based on infrared light filling according to
Picture quality carries out infrared light filling, the accuracy rate of night people's vehicle identification can be improved, simultaneously because using infrared light filling, to human eye
Without influence, light pollution will not be led to the problem of.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention, should
Understand, the present invention is not limited to implementation as described herein, the purpose of these implementations description is to help this field
In technical staff put into practice the present invention.Any those of skill in the art are easy to do not departing from spirit and scope of the invention
In the case of be further improved and perfect, therefore the present invention is only by the content of the claims in the present invention and limiting for range
System, intention, which covers, all to be included the alternative in the spirit and scope of the invention being defined by the appended claims and waits
Same scheme.
Claims (13)
1. people's vehicle recognition methods based on infrared light filling, which is characterized in that this method includes：
First step acquires scene image；
Second step calculates the luminance mean value of scene image, if luminance mean value is less than first threshold, opens infrared light filling,
And resurvey scene image；
Third step carries out vehicle detection and car plate detection to scene image, obtains vehicle region and license plate area；
Four steps carries out Car license recognition to license plate area, obtains license plate recognition result if there is license plate area；If no
There are license plate areas, then carry out vehicle cab recognition to vehicle region, obtain vehicle cab recognition result；
5th step, setting vehicle window region carry out Face datection to vehicle window region, obtain human face region, carried out to human face region
Identification obtains face recognition result；
6th step exports car plate or vehicle cab recognition result and face recognition result.
2. the value range of the method as described in claim 1, the first threshold is 90~110.
3. the method as described in claim 1, which is characterized in that the 5th step includes：
Vehicle window zone enactment steps, according to the position of license plate area, setting vehicle window region；If there is no license plate area, according to
The position of vehicle region, setting vehicle window region；
Face datection step obtains human face region using method for detecting human face from vehicle window region；
Human face region is compared with the human face data logged in database using face identification method for recognition of face step,
If compared unanimously, using the user information in database as the face information of human face region, if comparison is inconsistent, remember
Record is visitor, works out visitor in the database and numbers and store visitor's face information.
4. method as claimed in claim 3, which is characterized in that the vehicle window zone enactment steps include：
Vehicle window setting procedure based on car plate obtains left margin x=pl, the right margin x=of license plate area if there is license plate area
Pr, coboundary y=pt, lower boundary y=pb, set the left margin in vehicle window region asIt is right
Boundary isCoboundary is
Lower boundary isWpFor the width of license plate area, W is the width of scene image, 3 < of λ
λ2；
Vehicle window setting procedure based on vehicle obtains the upper subregion of vehicle region as vehicle if there is no license plate area
Window region.
5. method as claimed in claim 4, which is characterized in that the vehicle window setting procedure based on vehicle includes：If no
There are license plate areas, obtain coboundary y=vt, the lower boundary y=vb of vehicle region, the left margin in setting vehicle window region, the right
Boundary, coboundary are respectively the left margin, right margin, coboundary of vehicle region, set the lower boundary in vehicle window region as
6. method as claimed in claim 4,1 ∈ [ of the λ;1.2,1.8], 2 ∈ [ of λ;4.2,4.8], 3 ∈ [ of λ;0.3,0.8].
7. method as claimed in claim 3, further, the face identification method are the face based on convolutional neural networks
Recognition methods.
8. the method for claim 7, which is characterized in that the recognition of face step may include：It chooses marked
Face sample image, is trained convolutional neural networks, obtains trained face Network Recognition model；Using face network
Human face region is compared with the user's human face data logged in database for identification model, if compared unanimously, by data
Face information of the user information as human face region in library is recorded as visitor, compiles in the database if comparison is inconsistent
Visitor processed numbers and stores visitor's face information；
Wherein, the marked face sample image includes：Different angle, different brightness, different appearances under non-infrared light compensating lamp
The facial image for having marked classification of state, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
9. the method for claim 7, which is characterized in that the recognition of face step can also include：
First network identification model training step chooses marked Generic face sample image, to the first convolutional neural networks
It is trained, obtains trained first network identification model；
Second Network Recognition model classification step chooses marked infrared face sample image, to the second convolutional neural networks
It is trained, obtains trained second Network Recognition model；
Human face region identification step, when not opening infrared light filling, using first network identification model to human face region and data
The user's human face data logged in library is compared, when opening infrared light filling, using the second Network Recognition model pair and data
The user's human face data logged in library is compared；If compared unanimously, using the user information in database as face area
The face information in domain is recorded as visitor if comparison is inconsistent, works out visitor in the database and numbers and store visitor's face
Information；
Wherein, the marked Generic face sample image includes：Different angle under non-infrared light compensating lamp, different brightness, no
With the facial image for having marked classification of posture；The marked infrared face sample image includes：Under infrared light compensating lamp not
The facial image for having marked classification of same angle, different postures.
10. people's vehicle identification device based on infrared light filling, which is characterized in that the device includes：
Scene image acquisition module, for acquiring scene image；
Infrared light filling control module, the luminance mean value for calculating scene image are opened if luminance mean value is less than first threshold
Infrared light filling is opened, and resurveys scene image；
Vehicle license plate detection module obtains vehicle region and car plate for carrying out vehicle detection and car plate detection to scene image
Region；
Car plate vehicle cab recognition module, for if there is license plate area, carrying out Car license recognition to license plate area, obtaining Car license recognition
As a result；If there is no license plate area, then vehicle cab recognition is carried out to vehicle region, obtains vehicle cab recognition result；
Vehicle window region face recognition module carries out Face datection to vehicle window region, obtains face area for setting vehicle window region
Human face region is identified in domain, obtains face recognition result；
People's vehicle recognition result output module, for exporting car plate or vehicle cab recognition result and face recognition result.
11. device as claimed in claim 10, which is characterized in that vehicle window region face recognition module includes：Vehicle window area
Domain setting module, for the position according to license plate area, setting vehicle window region；If there is no license plate area, according to vehicle area
The position in domain, setting vehicle window region；
Face detection module obtains human face region for using method for detecting human face from vehicle window region；
Face recognition module carries out the human face data logged in human face region and database for using face identification method
It compares, if compared unanimously, using the user information in database as the face information of human face region, differs if compared
It causes, is then recorded as visitor, work out visitor in the database and number and store visitor's face information.
12. device as claimed in claim 11, which is characterized in that the face recognition module may include：For choosing
The face sample image of label, is trained convolutional neural networks, obtains trained face Network Recognition model；For adopting
Human face region is compared with the user's human face data logged in database with face Network Recognition model, if comparing one
It causes, then using the user information in database as the face information of human face region, if comparison is inconsistent, is recorded as visitor,
Establishment visitor numbers and stores visitor's face information in the database；
Wherein, the marked face sample image includes：Different angle, different brightness, different appearances under non-infrared light compensating lamp
The facial image for having marked classification of state, the facial image for having marked classification of different angle, different postures under infrared light compensating lamp.
13. device as claimed in claim 11, which is characterized in that the face recognition module can also include：
First network identification model training module, for choosing marked Generic face sample image, to the first convolutional Neural
Network is trained, and obtains trained first network identification model；
Second Network Recognition category of model module, for choosing marked infrared face sample image, to the second convolutional Neural
Network is trained, and obtains trained second Network Recognition model；
Human face region identification module, for when not opening infrared light filling, using first network identification model to human face region with
The user's human face data logged in database is compared, when opening infrared light filling, using the second Network Recognition model pair with
The user's human face data logged in database is compared；If consistent for comparing, the user information in database is made
It is recorded as visitor if comparison is inconsistent for the face information of human face region, visitor is worked out in the database and numbers and store
Visitor's face information；
Wherein, the marked Generic face sample image includes：Different angle under non-infrared light compensating lamp, different brightness, no
With the facial image for having marked classification of posture；The marked infrared face sample image includes：Under infrared light compensating lamp not
The facial image for having marked classification of same angle, different postures.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810495514.7A CN108710857B (en) | 2018-05-22 | 2018-05-22 | Method and device for identifying people and vehicles based on infrared supplementary lighting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810495514.7A CN108710857B (en) | 2018-05-22 | 2018-05-22 | Method and device for identifying people and vehicles based on infrared supplementary lighting |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108710857A true CN108710857A (en) | 2018-10-26 |
| CN108710857B CN108710857B (en) | 2022-05-17 |
Family
ID=63868389
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810495514.7A Active CN108710857B (en) | 2018-05-22 | 2018-05-22 | Method and device for identifying people and vehicles based on infrared supplementary lighting |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108710857B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109658527A (en) * | 2018-12-14 | 2019-04-19 | 广东安居宝数码科技股份有限公司 | Vehicle parking charge management method, device, equipment and storage medium |
| CN109766854A (en) * | 2019-01-15 | 2019-05-17 | 济南浪潮高新科技投资发展有限公司 | A kind of robust human face recognizer based on two stages complementary networks |
| CN109961026A (en) * | 2019-03-12 | 2019-07-02 | 天津瑟威兰斯科技有限公司 | Integrated license plate face recognition method and module based on raspberry group |
| CN110163089A (en) * | 2019-04-11 | 2019-08-23 | 西安艾润物联网技术服务有限责任公司 | Auth method and device |
| CN111160299A (en) * | 2019-12-31 | 2020-05-15 | 上海依图网络科技有限公司 | Living body identification method and device |
| CN112215234A (en) * | 2020-10-16 | 2021-01-12 | 神思电子技术股份有限公司 | Method for checking information of vehicle personnel at road interface |
| CN112861645A (en) * | 2021-01-18 | 2021-05-28 | 盛视科技股份有限公司 | Infrared camera dim light environment compensation method and device and electronic equipment |
| US11163977B2 (en) * | 2018-09-28 | 2021-11-02 | Panasonic I-Pro Sensing Solutions Co., Ltd. | Capturing camera |
| WO2022007633A1 (en) * | 2020-07-08 | 2022-01-13 | 北京大兴投资集团有限公司 | Entrance and exit management and control method |
| CN114187695A (en) * | 2021-12-08 | 2022-03-15 | 上海世茂物联网科技有限公司 | Method, device, system, equipment and medium for controlling barrier gate of garage |
| CN114255600A (en) * | 2022-01-26 | 2022-03-29 | 杭州海康威视数字技术股份有限公司 | Parking detection method, parking detection equipment and storage medium |
| CN114333329A (en) * | 2022-01-26 | 2022-04-12 | 杭州海康威视数字技术股份有限公司 | Parking detection method, parking detection equipment and storage medium |
Citations (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000149184A (en) * | 1998-11-10 | 2000-05-30 | Nippon Dainamatto Kk | Overloading control unit |
| CN103093528A (en) * | 2012-12-28 | 2013-05-08 | 中山大学 | Intelligent system for monitoring vehicles and people in community |
| CN103093517A (en) * | 2011-10-28 | 2013-05-08 | 台湾色彩与影像科技股份有限公司 | Human face recognition visitor management method |
| CN103544481A (en) * | 2013-10-24 | 2014-01-29 | 深圳市飞瑞斯科技有限公司 | Road gate monitoring method and system based on human face recognition |
| CN103781261A (en) * | 2014-02-28 | 2014-05-07 | 深圳英飞拓科技股份有限公司 | Infrared lamp control method for infrared network camera |
| CN104301623A (en) * | 2014-10-28 | 2015-01-21 | 浙江宇视科技有限公司 | Mixed light supplementing method and device |
| CN104317397A (en) * | 2014-10-14 | 2015-01-28 | 奇瑞汽车股份有限公司 | Vehicle-mounted man-machine interactive method |
| CN104966066A (en) * | 2015-06-26 | 2015-10-07 | 武汉大学 | Traffic block port monitoring oriented in-car human face detection method and system |
| CN105184873A (en) * | 2015-08-17 | 2015-12-23 | 深圳北斗应用技术研究院有限公司 | Intelligent parking lot management system |
| CN105430347A (en) * | 2015-11-30 | 2016-03-23 | 燕山大学 | Compressed sensing based wireless sensor image capturing and transmitting system |
| CN105530437A (en) * | 2016-01-29 | 2016-04-27 | 生迪智慧科技有限公司 | Environment brightness measurement and calculation method and shooting device |
| CN105844202A (en) * | 2015-01-12 | 2016-08-10 | 芋头科技（杭州）有限公司 | Image recognition system and method |
| CN106056071A (en) * | 2016-05-30 | 2016-10-26 | 北京智芯原动科技有限公司 | Method and device for detection of driver' behavior of making call |
| CN106485224A (en) * | 2016-10-13 | 2017-03-08 | 北京智芯原动科技有限公司 | A kind of seatbelt wearing recognition methodss and device |
| CN106503707A (en) * | 2016-10-21 | 2017-03-15 | 浙江宇视科技有限公司 | Licence plate recognition method and device under the conditions of a kind of infrared light filling |
| CN106909882A (en) * | 2017-01-16 | 2017-06-30 | 广东工业大学 | A kind of face identification system and method for being applied to security robot |
| CN107103903A (en) * | 2017-05-05 | 2017-08-29 | 百度在线网络技术（北京）有限公司 | Acoustic training model method, device and storage medium based on artificial intelligence |
| CN107248289A (en) * | 2017-07-01 | 2017-10-13 | 广东艾科智泊科技股份有限公司 | A kind of unidentified the self-service of car plate vehicle in many parking lots opens a sluice gate system |
| CN107301384A (en) * | 2017-06-09 | 2017-10-27 | 湖北天业云商网络科技有限公司 | A kind of driver takes phone behavioral value method and system |
| CN107370956A (en) * | 2017-08-21 | 2017-11-21 | 深圳市天视通电子科技有限公司 | A kind of method and starlight video camera for realizing the switching of starlight video camera diurnal pattern |
| CN107452104A (en) * | 2017-07-26 | 2017-12-08 | 北京声迅电子股份有限公司 | A kind of control method for vehicle and system of the vehicle bayonet socket based on intelligent monitoring |
| CN206741505U (en) * | 2017-03-17 | 2017-12-12 | 中控智慧科技股份有限公司 | Identify management equipment and device |
| CN107491750A (en) * | 2017-08-14 | 2017-12-19 | 深圳前海华夏智信数据科技有限公司 | A kind of testimony of a witness comparison method and device |
| CN107635099A (en) * | 2017-10-09 | 2018-01-26 | 深圳市天视通电子科技有限公司 | A kind of double optical-fiber network video cameras of human body sensing and safety defense monitoring system |
| US20180060648A1 (en) * | 2016-08-23 | 2018-03-01 | Samsung Electronics Co., Ltd. | Liveness test method and apparatus |
| WO2018076523A1 (en) * | 2016-10-25 | 2018-05-03 | 科世达（上海）管理有限公司 | Gesture recognition method and apparatus, and in-vehicle system |
-
2018
- 2018-05-22 CN CN201810495514.7A patent/CN108710857B/en active Active
Patent Citations (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2000149184A (en) * | 1998-11-10 | 2000-05-30 | Nippon Dainamatto Kk | Overloading control unit |
| CN103093517A (en) * | 2011-10-28 | 2013-05-08 | 台湾色彩与影像科技股份有限公司 | Human face recognition visitor management method |
| CN103093528A (en) * | 2012-12-28 | 2013-05-08 | 中山大学 | Intelligent system for monitoring vehicles and people in community |
| CN103544481A (en) * | 2013-10-24 | 2014-01-29 | 深圳市飞瑞斯科技有限公司 | Road gate monitoring method and system based on human face recognition |
| CN103781261A (en) * | 2014-02-28 | 2014-05-07 | 深圳英飞拓科技股份有限公司 | Infrared lamp control method for infrared network camera |
| CN104317397A (en) * | 2014-10-14 | 2015-01-28 | 奇瑞汽车股份有限公司 | Vehicle-mounted man-machine interactive method |
| CN104301623A (en) * | 2014-10-28 | 2015-01-21 | 浙江宇视科技有限公司 | Mixed light supplementing method and device |
| CN105844202A (en) * | 2015-01-12 | 2016-08-10 | 芋头科技（杭州）有限公司 | Image recognition system and method |
| CN104966066A (en) * | 2015-06-26 | 2015-10-07 | 武汉大学 | Traffic block port monitoring oriented in-car human face detection method and system |
| CN105184873A (en) * | 2015-08-17 | 2015-12-23 | 深圳北斗应用技术研究院有限公司 | Intelligent parking lot management system |
| CN105430347A (en) * | 2015-11-30 | 2016-03-23 | 燕山大学 | Compressed sensing based wireless sensor image capturing and transmitting system |
| CN105530437A (en) * | 2016-01-29 | 2016-04-27 | 生迪智慧科技有限公司 | Environment brightness measurement and calculation method and shooting device |
| CN106056071A (en) * | 2016-05-30 | 2016-10-26 | 北京智芯原动科技有限公司 | Method and device for detection of driver' behavior of making call |
| US20180060648A1 (en) * | 2016-08-23 | 2018-03-01 | Samsung Electronics Co., Ltd. | Liveness test method and apparatus |
| CN106485224A (en) * | 2016-10-13 | 2017-03-08 | 北京智芯原动科技有限公司 | A kind of seatbelt wearing recognition methodss and device |
| CN106503707A (en) * | 2016-10-21 | 2017-03-15 | 浙江宇视科技有限公司 | Licence plate recognition method and device under the conditions of a kind of infrared light filling |
| WO2018076523A1 (en) * | 2016-10-25 | 2018-05-03 | 科世达（上海）管理有限公司 | Gesture recognition method and apparatus, and in-vehicle system |
| CN106909882A (en) * | 2017-01-16 | 2017-06-30 | 广东工业大学 | A kind of face identification system and method for being applied to security robot |
| CN206741505U (en) * | 2017-03-17 | 2017-12-12 | 中控智慧科技股份有限公司 | Identify management equipment and device |
| CN107103903A (en) * | 2017-05-05 | 2017-08-29 | 百度在线网络技术（北京）有限公司 | Acoustic training model method, device and storage medium based on artificial intelligence |
| CN107301384A (en) * | 2017-06-09 | 2017-10-27 | 湖北天业云商网络科技有限公司 | A kind of driver takes phone behavioral value method and system |
| CN107248289A (en) * | 2017-07-01 | 2017-10-13 | 广东艾科智泊科技股份有限公司 | A kind of unidentified the self-service of car plate vehicle in many parking lots opens a sluice gate system |
| CN107452104A (en) * | 2017-07-26 | 2017-12-08 | 北京声迅电子股份有限公司 | A kind of control method for vehicle and system of the vehicle bayonet socket based on intelligent monitoring |
| CN107491750A (en) * | 2017-08-14 | 2017-12-19 | 深圳前海华夏智信数据科技有限公司 | A kind of testimony of a witness comparison method and device |
| CN107370956A (en) * | 2017-08-21 | 2017-11-21 | 深圳市天视通电子科技有限公司 | A kind of method and starlight video camera for realizing the switching of starlight video camera diurnal pattern |
| CN107635099A (en) * | 2017-10-09 | 2018-01-26 | 深圳市天视通电子科技有限公司 | A kind of double optical-fiber network video cameras of human body sensing and safety defense monitoring system |
Non-Patent Citations (4)
| Title |
|---|
| KAIGUO 等: "Face recognition using both visible light image and near-infrared image and a deep network", 《CAAI TRANSACTIONS ON INTELLIGENCE TECHNOLOGY》 * |
| 曹永峰 等: "基于GA-BP神经网络的计算机智能化图像识别技术探究", 《应用激光》 * |
| 王永强 等: "基于图像识别技术的车辆管理系统的改进研究", 《现代制造技术与装备》 * |
| 葛如海: "基于GA-BP的安全带佩戴识别方法", 《江苏大学学报》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11163977B2 (en) * | 2018-09-28 | 2021-11-02 | Panasonic I-Pro Sensing Solutions Co., Ltd. | Capturing camera |
| CN109658527A (en) * | 2018-12-14 | 2019-04-19 | 广东安居宝数码科技股份有限公司 | Vehicle parking charge management method, device, equipment and storage medium |
| CN109766854A (en) * | 2019-01-15 | 2019-05-17 | 济南浪潮高新科技投资发展有限公司 | A kind of robust human face recognizer based on two stages complementary networks |
| CN109961026A (en) * | 2019-03-12 | 2019-07-02 | 天津瑟威兰斯科技有限公司 | Integrated license plate face recognition method and module based on raspberry group |
| CN110163089A (en) * | 2019-04-11 | 2019-08-23 | 西安艾润物联网技术服务有限责任公司 | Auth method and device |
| CN111160299A (en) * | 2019-12-31 | 2020-05-15 | 上海依图网络科技有限公司 | Living body identification method and device |
| WO2022007633A1 (en) * | 2020-07-08 | 2022-01-13 | 北京大兴投资集团有限公司 | Entrance and exit management and control method |
| CN112215234A (en) * | 2020-10-16 | 2021-01-12 | 神思电子技术股份有限公司 | Method for checking information of vehicle personnel at road interface |
| CN112861645A (en) * | 2021-01-18 | 2021-05-28 | 盛视科技股份有限公司 | Infrared camera dim light environment compensation method and device and electronic equipment |
| CN114187695A (en) * | 2021-12-08 | 2022-03-15 | 上海世茂物联网科技有限公司 | Method, device, system, equipment and medium for controlling barrier gate of garage |
| CN114187695B (en) * | 2021-12-08 | 2023-12-08 | 上海世茂物联网科技有限公司 | Barrier gate control method, device, system, equipment and medium for garage |
| CN114255600A (en) * | 2022-01-26 | 2022-03-29 | 杭州海康威视数字技术股份有限公司 | Parking detection method, parking detection equipment and storage medium |
| CN114333329A (en) * | 2022-01-26 | 2022-04-12 | 杭州海康威视数字技术股份有限公司 | Parking detection method, parking detection equipment and storage medium |
| CN114333329B (en) * | 2022-01-26 | 2023-09-29 | 杭州海康威视数字技术股份有限公司 | Parking detection method, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108710857B (en) | 2022-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108710857A (en) | People's vehicle recognition methods based on infrared light filling and device | |
| CN113256965B (en) | Traffic risk sensing system and method based on smart city | |
| CN110298278B (en) | Underground parking garage pedestrian and vehicle monitoring method based on artificial intelligence | |
| CN101344988B (en) | Image acquisition and processing equipment and method, vehicle monitoring and recording system | |
| CN105654733B (en) | A kind of front and back licence plate recognition method and device based on video detection | |
| CN106372576A (en) | Deep learning-based intelligent indoor intrusion detection method and system | |
| CN102509090B (en) | A kind of vehicle feature recognition device based on public safety video image in sky net engineering | |
| CN106022243B (en) | A kind of retrograde recognition methods of the car lane vehicle based on image procossing | |
| CN1920896A (en) | Real-time image monitoring and genuine-fake identification system for vehicle with automatic color regulating and light filling license plate | |
| CN105654734A (en) | Data processing method and device for vehicle violations | |
| CN109671189B (en) | Intelligent access system based on human face and vehicle associated characteristics | |
| CN105160297A (en) | Masked man event automatic detection method based on skin color characteristics | |
| CN104102921A (en) | Video detection method and video detection device based on face recognition | |
| KR20150134548A (en) | Apparatus for vehicle detection | |
| CN101388146A (en) | Image acquisition and treatment apparatus and method, and vehicle monitoring and recording system | |
| KR101218302B1 (en) | Method for location estimation of vehicle number plate | |
| KR101151523B1 (en) | apparatus for recognizing number of vehicle using dual camera and method thereof | |
| KR101870641B1 (en) | Image surveillance system and method thereof | |
| CN111950399A (en) | Intelligent integrated equipment for face recognition and monitoring of computer room | |
| KR102309106B1 (en) | Situation linkage type image analysis device | |
| CN109817009A (en) | A method of obtaining unmanned required dynamic information | |
| KR102171384B1 (en) | Object recognition system and method using image correction filter | |
| CN105206060B (en) | A kind of vehicle type recognition device and its method based on SIFT feature | |
| CN111724605A (en) | Black smoke vehicle monitoring system and monitoring method thereof | |
| US10694112B2 (en) | Vehicular image pickup device and image capturing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |