CN110458985A - A kind of engineering site in-out management system and method - Google Patents
A kind of engineering site in-out management system and method Download PDFInfo
- Publication number
- CN110458985A CN110458985A CN201910666189.0A CN201910666189A CN110458985A CN 110458985 A CN110458985 A CN 110458985A CN 201910666189 A CN201910666189 A CN 201910666189A CN 110458985 A CN110458985 A CN 110458985A
- Authority
- CN
- China
- Prior art keywords
- data
- face
- image
- alarm
- matching
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
A kind of engineering site in-out management system of disclosure of the invention and method, including data acquisition facility, data prediction device, cloud server and warning device, data acquisition facility is used to obtain the record data of vehicle or personnel's entry/exit engineering site, and the record data are uploaded into data prediction device, data prediction device is for pre-processing the record data, and it uploads in cloud server, the cloud server is used for pretreated record data and preset record attribute Data Matching, by record matching object command returned data pretreatment unit, warning device is used to obtain data prediction device and issues the record data result instruction that it fails to match, to alarm;Can direct noninductive passage, avoid entering the defect of engineering site using card or artificial let pass, while image data is uploaded in cloud server and stored, ensured the authenticity of engineering truck disengaging site record.
Description
Technical field
The present invention relates to engineering site management domain more particularly to a kind of engineering site in-out management system and methods.
Background technique
Building Trade in China is quickly grown in recent years, is needed to introduce a large amount of workers and is carried out operation, but since worker is generally literary
Change degree is lower, and mobility is larger, difficult management, and on construction site, there is only various heavy mechanical equipments, and also there are many build
Material is built, therefore is the ground that takes place frequently of safety accident, only this needs the vehicle to engineering site, personnel to be monitored, to reduce peace
A possibility that full accident occurs.
When conventional truck or personnel enter engineering site, to the vehicle or staff for then having allotted disengaging card, management
Mode is to enter building site by swiping the card, i.e., driver or staff, which actively swipe the card before card reader, can enter building site, but this
There is very big drawback in the mode of sample.For the construction site of relatively large-scale, since worker is constantly in high frequency operation, often occur
It carries card to lose, but cannot reissue in time, cause worker impassable;Card facilitates worker arbitrarily to hand over due to easy to carry
It changes, will occur in this way for phenomenon of checking card, and traditional access control system can not specification visitor, administrative staff, construction personnel
Time and permission Deng disengaging building site.
When passing in and out engineering site for the vehicle or personnel of not allotting disengaging card, by letting pass after manually verifying or forbidding leading to
Row, not only manually veritifies at high cost, but also makes the vehicle/personnel for importing and exporting or delimiting disengaging region mouthful in engineering site current
Time-consuming, low efficiency.
Summary of the invention
Technical problems based on background technology, the invention proposes a kind of engineering site in-out management system and sides
Method carries out noninductive detection to the vehicle of disengaging engineering, personnel by existing, realizes to illegally into the vehicle of engineering, the announcement of personnel
It is alert.
A kind of engineering site in-out management system proposed by the present invention, including data acquisition facility, data prediction device,
Cloud server and warning device;
Data acquisition facility is used to obtain the record data of vehicle or personnel's entry/exit engineering site, and by the record number
According to data prediction device is uploaded to, the record data include license plate image data and driver's image data;
Data prediction device uploads in cloud server for pre-processing to the record data;
The cloud server is used for pretreated record data and preset record attribute Data Matching, will
Record matching object command returned data pretreatment unit;
Warning device is used to obtain data prediction device and issues the record data result instruction that it fails to match, to carry out
Alarm.
Further, the data prediction device includes for carrying out the pretreated facial image of recognition of face to personnel
Processing module, the face image processing module include luminous intensity computing unit, luminous intensity comparing unit, the first recognition of face list
Member and the second face identification unit, the luminous intensity computing unit includes luminous intensity conversion subunit and light intensity meter operator list
Member;
The luminous intensity computing unit obtains the light intensity of current environment for detecting to the luminous intensity of current environment
Degree;
Luminous intensity comparing unit is for the luminous intensity of current environment to be compared with the Intensity threshold of setting;
First face identification unit is used to utilize when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting
The RGB image that data acquisition facility uploads carries out recognition of face, and the facial image after identification is uploaded to cloud server and is carried out
Matching verifying, when matching authentication failed, issues alarm command to warning device;
Second face identification unit is used to utilize the number when the luminous intensity of current environment is less than the intensity threshold of setting
The RGB image and depth image uploaded according to acquisition device carries out recognition of face, and the facial image after identification is uploaded to cloud clothes
Business device carries out matching verifying, when matching authentication failed, issues alarm command to warning device;
The luminous intensity conversion subunit is used to the RGB image being converted into HSI image, and extracts the image in the channel I;
The luminous intensity computation subunit be used to be calculated according to the image in the channel I the average value of the image in the channel I and by its
Light intensity value as current environment.
Further, the first face identification unit includes the first training set subelement, first extracts subelement and the first
Face coupling subelement, second face identification unit include the first mapping subelement, the second extraction subelement, the second mapping
Unit and the second face coupling subelement;
The first training set subelement is used to adopt in advance when ambient light intensity is greater than or equal to the intensity threshold of setting
The RGB image and depth image for collecting face constitute the first training dataset, and carry out feature to the RGB image in the first training set
It extracts, obtained first eigenvector is stored to database;
Described first, which extracts subelement, is used for when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, right
The face RGB image acquired in real time carries out feature extraction, obtains feature vector;
First face coupling subelement is for carrying out the first eigenvector stored in this feature vector and database
Match, recognition of face is carried out, and send the result of recognition of face matching authentication failed in warning device, to alarm;
First mapping subelement is used to acquire face in advance when the luminous intensity of environment is less than the intensity threshold of setting
RGB image and depth image, and the region on the RGB image where face is extracted, and utilize mapping equation by the RGB image
It is mapped on the depth image, obtains three dimensional point cloud to construct the second training dataset;
The three dimensional point cloud that second extraction subelement is used to concentrate the second training data carries out feature extraction, and will mention
The second feature vector got is stored to database;
Second mapping subelement is used to utilize mapping equation when the luminous intensity of current environment is less than the intensity threshold of setting
The RGB image and depth image of the face acquired in real time are handled, real-time three-dimensional point cloud data is obtained;
Second face coupling subelement is used for the feature that real-time three-dimensional point cloud data feature is extracted, and will be extracted
Vector is matched with the second feature vector in database, carries out recognition of face, and recognition of face is matched authentication failed
As a result it is sent in warning device, to alarm.
The mapping equation is
Wherein, (u, v) is the coordinate value of corresponding points on the RGB image, IrIt is color camera internal reference, T is RGB camera shooting
The module and carriage transformation matrix p=(x, y, z, 1) of [R t] between head coordinate system and depth camera coordinate systemT, indicate that three-dimensional point is sat
Mark, * indicate product.
Further, the cloud server include authority checking module, compare interface module, query interface module, on
Pass interface module and block chain storage module;
Authority checking module is used for the checking request sent according to the data prediction device, to locate in advance to the data
It manages device and carries out authority checking;
Interface module is compared for data prediction will to be back to by the verification result of authority checking module authority checking
Device;
Query interface module is used for after the authority checking module verification passes through, and calls data verification interface to the symbol
The automobile ingress/egress record data for closing inquiry instruction are verified, and are verified as a result, and verification result is back to the service
Device end;
Upload interface module be used for after authority checking module verification passes through for the data prediction device call with incite somebody to action
The automobile ingress/egress record data are uploaded to block chain storage module.
Further, the data acquisition facility includes image collecting device and connect with image collecting device output end
Image conversion apparatus;
Described image acquisition device includes to the first camera identified by gate entry/exit vehicle, for logical
The driver personnel for crossing gate entry/exit vehicle carry out the second camera of recognition of face and for crossing gate entry/exit to non-through
Vehicle or the third camera that is identified of personnel, the output end of the first camera, second camera and third camera
It is all connected to the input terminal of image conversion apparatus;
Image conversion apparatus includes video decoding circuit and is filtered the first filter circuit of noise reduction to image, the first filter
The output end of wave circuit and the input terminal of video decoding circuit connect;
Video decoding circuit includes video decoding chip, crystal oscillating circuit and the second filter circuit, the output end of crystal oscillating circuit
It is connected to the input terminal of video decoding chip, the input terminal of the second filter circuit is connected to the output end of external power supply, filtered electrical
The output end on road is connected to the analogue ground of frequency decoding chip;
First filter circuit includes capacitor and two resistance, wherein a resistance one end be connected to interface P1 output end,
The other end is connected respectively to capacitance cathode end and another resistance one end, the other end ground connection of another resistance, the connection of capacitor negative pole end
Onto the input pin of the video decoding chip;
Second filter circuit includes at least one filter capacitor, is connected in parallel to each other between filter capacitor, after parallel connection just
It is extremely connected to external power supply end, the negative pole end after parallel connection is connected to the analogue ground pin of the video decoding chip.
Further, warning device includes linkage matching unit and interlink warning unit, the linkage matching unit include
Unauthorized passes in and out alarm part-unit, the alarm part-unit that do not wear a safety helmet, abnormal time disengaging alarm part-unit, non-same people's disengaging
Alarm part-unit;
Link matching unit be used for based on pre-set alarm parameters and linkage parameter matching relationship, obtain with currently
The corresponding linkage characteristic parameter of alarm parameters, current alerts parameter are the alarm commands that the data prediction device issues;
Interlink warning unit is sent in warning device corresponding with linkage characteristic parameter for the characteristic parameter that will link, with
Realize early warning linkage;
Unauthorized disengaging alarm part-unit be used to judge the type of alarm of acquired alert data whether be unauthorized into
It alarms out, if so, linkage characteristic parameter is sent in corresponding warning device with the first alert levels and carries out interlink warning,
If it is not, then entering the alarm part-unit that do not wear a safety helmet;
The alarm part-unit that do not wear a safety helmet is for judging whether type of alarm is the alarm that do not wear a safety helmet;If so, linkage
Characteristic parameter is sent in corresponding warning device with the second alert levels and carries out interlink warning, if it is not, when then entering abnormal
Between pass in and out alarm part-unit;
Abnormal time disengaging alarm part-unit is used to judge whether type of alarm to be abnormal time disengaging alarm, if so,
Linkage characteristic parameter is sent in corresponding warning device with third alert levels carries out interlink warning, if it is not, then entering non-
Same people passes in and out alarm part-unit;
Non- same people passes in and out alarm part-unit and is used to judge whether the type of alarm of acquired alert data to be non-same
People passes in and out alarm, if so, linkage characteristic parameter is sent in corresponding warning device with the 4th alert levels and is linked
Alarm, if it is not, then without alarm.
A kind of engineering site disengaging management method, in data prediction device, comprising:
The record data for obtaining vehicle or personnel's entry/exit engineering site are pre-processed, and upload to cloud server
In;
The record matching object command that cloud server returns is obtained, the record matching object command is pretreated
Record the output result in data and cloud server after preset record attribute Data Matching;
It sends the record matching object command that it fails to match in warning device, to alarm.
Further, when being pre-processed to the record data for obtaining vehicle or personnel's entry/exit engineering site, when to people
When face carries out identification matching, comprising:
The RGB image and depth image of face are obtained respectively;
The luminous intensity of current environment is detected, the luminous intensity of current environment is obtained;
The luminous intensity of current environment is compared with the Intensity threshold of setting;
When the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, face is carried out using the RGB image
Identification, uploads to cloud server for the facial image after identification and carries out matching verifying, when matching authentication failed, fills to alarm
It sets and issues alarm command;
When the luminous intensity of current environment is less than the intensity threshold of setting, carried out using the RGB image and depth image
Facial image after identification is uploaded to cloud server and carries out matching verifying by recognition of face, when matching authentication failed, to report
Alarm device issues alarm command.
A kind of engineering site disengaging management method, in cloud server, comprising:
Record data after obtaining data prediction, the pretreated record data are pretreatment units to acquired
Record data carry out pretreated data;
It, will be under record matching object command by pretreated record data and preset record attribute Data Matching
It is sent to pretreatment unit, the instruction that it fails to match in the record matching object command is for triggering warning device to be reported
It is alert;
Further, when being pre-processed to the record data for obtaining vehicle or personnel's entry/exit engineering site, when to people
When face carries out identification matching, comprising:
The RGB image and depth image of face are obtained respectively;
The luminous intensity of current environment is detected, the luminous intensity of current environment is obtained;
The luminous intensity of current environment is compared with the Intensity threshold of setting;
When the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, face is carried out using the RGB image
Identification, uploads to cloud server for the facial image after identification and carries out matching verifying, when matching authentication failed, fills to alarm
It sets and issues alarm command;
When the luminous intensity of current environment is less than the intensity threshold of setting, carried out using the RGB image and depth image
Facial image after identification is uploaded to cloud server and carries out matching verifying by recognition of face, when matching authentication failed, to report
Alarm device issues alarm command.
The advantages of a kind of engineering site in-out management system provided by the invention and method, is: providing in structure of the invention
A kind of engineering site in-out management system and method, by acquired disengaging engineering site vehicle/personnel record data
It is matched with the preset record attribute data of cloud server, is sent the object command that it fails to match in warning device,
To be alarmed accordingly, while the vehicle of successful match or personnel are to direct noninductive passage, avoid using card or artificial
It lets pass and enters the defect of engineering site;Cloud server stores the image of the successful match uploaded every time, has ensured work
The authenticity of journey vehicles while passing site record, not only ensure that the convenient and efficient of data application, but also ensure that the safety of data is true
Property, be conducive to the requirement of standardization engineering truck disengaging site record data, ensure the integrality of record data management;Light
When preferable, data prediction device directly use data acquisition facility upload RGB image carry out 2D image recognition, light compared with
When poor, data prediction device carries out Face datection using the RGB image that data acquisition facility uploads, and then extracts corresponding people
Face three-dimensional point cloud carries out recognition of face, i.e., 2D image is only to provide detection face effect, and identifies without 2D, using 3D point cloud
Carry out recognition of face, improve the stability and accuracy of the identification of different light intensity human faces, so improve to exception into
The accurate alarm of engineering site out;Warning device is obtained according to the type of alarm being previously set and the corresponding relationship of alert levels
Corresponding alert levels, and then the characteristic parameter that links accordingly is obtained, realize corresponding early warning linkage under alert data.
Detailed description of the invention
Fig. 1 is a kind of functional block diagram of engineering site in-out management system of the present invention;
Fig. 2 is the structural schematic diagram of image acquiring device of the present invention;
Fig. 3 is the circuit topology figure of image conversion apparatus of the present invention;
Fig. 4 is that a kind of engineering site of the present invention passes in and out management method, for the step schematic diagram in data prediction device;
Fig. 5 is that a kind of engineering site of the present invention passes in and out management method, for the step schematic diagram in cloud server;
Fig. 6 is the detailed step flow chart of this hair step S0003.
Wherein, 1- data acquisition facility, 2- data prediction device, 3- cloud server, 4- warning device, 11- image
Acquisition device, 12- image conversion apparatus, 20- human face data pretreatment unit, 21- license plate data prediction device, 111- first
Camera, 112- second camera, 113- third camera, 121- video decoding circuit, the first filter circuit of 122-, when 123-
Clock control interface, 201- luminous intensity computing unit, 202- luminous intensity comparing unit, the first face identification unit of 203-, 204-
Two face identification units, 31- authority checking module, 32- compare interface module, 33- query interface module, and 34- uploads interface mould
Block, 35- block chain storage module, 41- linkage matching unit, 42- interlink warning unit, 411- unauthorized disengaging alarm are single
Member, 412- do not wear a safety helmet alarm part-unit, and 413- abnormal time passes in and out alarm part-unit, and the non-same people of 414- passes in and out alarm
Unit.
Specific embodiment
In the following, technical solution of the present invention is described in detail by specific embodiment, illustrate in the following description
Many details are to fully understand the present invention.But the present invention can be to be much different from its other party described herein
Formula is implemented, and those skilled in the art can make similar improvements without departing from the spirit of the invention, therefore the present invention is not
It is limited by following public specific implementation.
Referring to Fig.1, a kind of engineering site in-out management system proposed by the present invention and method, including data acquisition facility,
Data prediction device and warning device;
Data acquisition facility 1 is used to obtain the record data of vehicle or personnel's entry/exit engineering site, and by the record number
According to uploading to data prediction device 2;
Data prediction device 2 uploads in cloud server 3 for pre-processing to the record data;
Cloud server 3 is used to remember pretreated record data and preset record attribute Data Matching
It records matching result and instructs returned data pretreatment unit 2;
Warning device 4 is used to obtain data prediction device 2 and issues the instruction of the record data result that it fails to match, with into
Row alarm.
It is acquired by the image data to disengaging engineering site vehicle, personnel and uploads to data prediction device 2
It after being pre-processed, uploads to cloud server 3 and carries out images match verifying, whether to be carried out the alarm of illegal entry/exit
Signal carries out license plate to vehicle simultaneously and driver identifies, when detecting that any one is illegal entry/exit, then alarms, when
When detecting that license plate and driver mismatch, also alarms, engineering truck is avoided to be driven by unauthorized person, cause engineering safety
The harm of accident.
As shown in Figures 1 to 3, the data acquisition facility 1 includes image collecting device 11 and exports with image collecting device
Hold the image conversion apparatus 12 of connection;
Described image acquisition device 11 includes to the first camera 111 identified by gate entry/exit vehicle, uses
In the second camera 112 to the driver personnel progress recognition of face by gate entry/exit vehicle and for passing through to non-
The third camera 113 that the vehicle of gate entry/exit or personnel are identified, the first camera 111, second camera 112 and
The output end of third camera 113 is all connected to the input terminal of image conversion apparatus, the first camera 111, second camera 112
The image that will acquire is sent in image conversion apparatus, and the output end of third camera 113 also connects with the input terminal of warning device
It connects, when third camera 113 detects the non-through vehicle for crossing gate entry/exit or personnel, on the one hand directly passes through warning device
It alarms, on the other hand sends the image accordingly obtained in image conversion apparatus;First camera 111 is fixed on first
On support rod, first support bar is fixed at gate, and first support bar uses existing support rod, is taken the photograph with that can stablize fixation
Picture head is preferable fixed link and stock fixed form, and second camera 112 is set in engineering site and away from a spacing at gate
From L, the distance of distance L is not strictly limited, and is retrieved as preferable distance to be able to achieve the stabilization of image, in the present embodiment preferably
L is 2-5m, the first camera 111, second camera 112, third camera 113 height can effectively obtain corresponding image
Subject to, be typically provided to 2-6m, the first camera 111, second camera 112, third camera 113 mounting means have it is more
Kind, for example, installed by pre-set support rod, by being installed on gate, by being installed on temporary structure
Deng, the first camera 111 with can effectively detect the license plate image of vehicle for preferable mounting means, second camera 112 with
The facial image that can effectively detect the driver of vehicle is preferable mounting means, and third camera 113 is can effectively detect
Image to the non-through vehicle image for crossing gate entry/exit or personnel is preferable mounting means.
Image conversion apparatus 12 includes video decoding circuit 121 and the first filter circuit that noise reduction is filtered to image
122, the output end of the first filter circuit 122 is connect with the input terminal of video decoding circuit 121;
Video decoding circuit 121 includes video decoding chip, crystal oscillating circuit and the second filter circuit, the output of crystal oscillating circuit
End is connected to the input terminal of video decoding chip, and the input terminal of the second filter circuit is connected to the output end of external power supply, filters
The output end of circuit is connected to the analogue ground of frequency decoding chip;
Second filter circuit includes at least one filter capacitor, is connected in parallel to each other between filter capacitor, after parallel connection just
It is extremely connected to external power supply end, the negative pole end after parallel connection is connected to the analogue ground pin of the video decoding chip.At this
In embodiment, the capacitor in the second filter circuit may be 1 extremely multiple, and the present embodiment is for 4, the second filter circuit
Including capacitor C7, capacitor C8 and capacitor C9, capacitor C7, capacitor C8, be connected in parallel to each other between capacitor C9, it is in parallel after positive terminal connection
To the end external power supply VCC, the negative pole end after parallel connection is connected to the analogue ground pin of the video decoding chip.Pass through the second filter
Wave circuit makes the input/output waveform of chip SAA7111 voltage more smooth.
As shown in figure 4, video decoding chip uses the chip of model SAA7111A, first filter circuit 122 is wrapped
Capacitor and two resistance are included, wherein a resistance one end is connected to the output end of interface P1, the other end is connected respectively to capacitance cathode end
It is grounded with the other end of another resistance one end, another resistance, the input that capacitor negative pole end is connected to the video decoding chip is drawn
On foot;Multiple first filter circuits 122 can be set to carry out independent noise reduction filtering processing to the unlike signal of input.In this reality
It applies in example, the first filter circuit 122 includes for be filtered at gate by the vehicle image of gate entry/exit
One sub-circuit, the second sub-circuit for being filtered to driver's facial image by gate entry/exit vehicle and
For the third sub-circuit that the image to the non-through vehicle for crossing gate entry/exit or personnel is filtered, one can also be arranged
Extend the 4th spare sub-circuit;
First sub-circuit includes resistance R1, resistance R5, capacitor C1, and the one end resistance R1 is connected to 1 pin of terminals P 1, electricity
Resistance one end R1 is connected respectively to the positive terminal of one end of resistance R5, capacitor C1, and the other end ground connection of resistance R5, capacitor C1's is negative
Extremely it is connected on the AI21 pin of chip SAA7111;Second sub-circuit includes that the 4th analog to digital conversion circuit includes resistance R2, electricity
R6, capacitor C2 are hindered, the one end resistance R2 is connected to 1 pin of terminals P 2, and the one end resistance R2 is connected respectively to one end of resistance R6, electricity
Hold the positive terminal of C2, the other end ground connection of resistance R6, the negative pole end of capacitor C2 is connected on the AI22 pin of chip SAA7111;
Third sub-circuit resistance R3, resistance R7, capacitor C3, the one end resistance R3 are connected to 1 pin of terminals P 3, the one end resistance R3 difference
It is connected to one end of resistance R7, the positive terminal of capacitor C3, the other end ground connection of resistance R7, the negative pole end of capacitor C3 is connected to core
On the AI11 pin of piece SAA7111;4th sub-circuit includes resistance R4, resistance R8, capacitor C4, and the one end resistance R4 is connected to terminal
1 pin of P4, the one end resistance R4 are connected respectively to the positive terminal of one end of resistance R8, capacitor C4, and the other end of resistance R8 is grounded,
The negative pole end of capacitor C4 is connected on the AI12 pin of SAA7111 chip;Pass through above respectively arranged first filter circuit
Each sub-circuit carries out independent filtering noise reduction process to different target images, and the digital-to-analogue for being then linked into SAA7111 chip turns
It changes on pin and analog signal is converted to digital signal, the target image is then uploaded into cloud again, reduces target image
Transmission quantity, ensure that data transmission speed and real-time.
As shown in figure 4, SAA7111 chip passes through work clock needed for crystal oscillating circuit generation system, the crystal oscillating circuit
Including with capacitor C11, the both ends crystal Y7 are connected respectively to the input terminal of the video decoding chip, and the positive terminal of C11 connects respectively
It is connected to the both ends of crystal Y7, the negative pole end ground connection of C11.In the present embodiment, the both ends of crystal Y7 are connected respectively to chip
On XTAL, XTALI pin of SAA7111, the LLC pin of SAA7111 chip is connected to clock control interface by resistance R9
On 123, third filter circuit, third filter circuit packet are connected between clock control interface 123 and the video decoding chip
Include resistance R10 and resistance R11, one end of resistance R10 and resistance R11 are all connected to the output end of external power supply, resistance R10's
The other end is connected on the serial clock input pin (SCL pin) of the video decoding chip, and the other end of resistance R11 connects
It is connected on serial date transfer/output pin (SDA pin) of the video decoding chip.SAA7111 chip passes through crystal oscillator electricity
Road generates row locked clock signal at LLC pin, so that the clock control interface 123 generates periodic signal, needs to illustrate
, control of the periodic signal by SAA7111 chip CE pin level, when SAA7111 chip CE pin is high level,
Row locked clock signal is generated at SAA7111 chip LLC pin, so that clock control interface 123 generates periodic signal, when
When SAA7111 chip CE pin is low level, row locked clock signal, and then clock are not generated at SAA7111 chip LLC pin
Control interface 123 does not generate periodic signal.By the crystal oscillating circuit external clock device, with to handling number on chip SAA7111
According to when carry out periodic signal triggering control, with guarantee data transmission real-time.
For data acquisition facility 1 before uploading record data to data prediction device 2, which includes structuring number
According to, image data, the data structure of structural data include: information of vehicles (brand number, number plate type, type of vehicle etc.),
Pass in and out time, disengaging position, Image Name, driver etc.;Image data, that is, vehicles while passing is taken pictures picture and video.To record
Whether data format, which meets the requirements, is verified, and key step is as follows:
(1) judge whether information of vehicles is complete, information of vehicles includes at least brand number and returns to mistake if imperfect
Information;
(2) if information of vehicles is complete, continue to judge to pass in and out the time, disengaging position, driver's information whether be not
Sky, if the item of promising sky, returns to error message;
(3) judge whether vehicles while passing image information is complete, image information includes at least a vehicles while passing image, if not
Completely, then error message is returned.
As shown in Figure 1, data prediction device 2 includes pre- for carrying out the pretreated human face data of recognition of face to personnel
Processing unit 20 and for license plate carry out image preprocessing license plate data prediction device 21, license plate data prediction device
21 can carry out image preprocessing to license plate using existing licence plate recognition method, and pretreated license plate image is uploaded to
In cloud server, to carry out license plate matching verifying, when being verified, lets pass to vehicle, when authentication failed, pass through report
Alarm device is alarmed, and is identified, is specifically included to face by RGB-D camera in the present embodiment:
As shown in Figure 1, the human face data pretreatment unit 20 is more single including luminous intensity computing unit 201, light strength ratio
First 202, first face identification unit 203 and the second face identification unit 204;
The luminous intensity computing unit 201 obtains the light of current environment for detecting to the luminous intensity of current environment
Intensity;
Luminous intensity comparing unit 202 is for the luminous intensity of current environment to be compared with the Intensity threshold of setting;
First face identification unit 203 is used for when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting,
Recognition of face is carried out using the RGB image that data acquisition facility uploads, the facial image after identification is uploaded into cloud server
Matching verifying is carried out, when matching authentication failed, issues alarm command to warning device;
Second face identification unit 204 is used to utilize institute when the luminous intensity of current environment is less than the intensity threshold of setting
The RGB image and depth image for stating data acquisition facility upload carry out recognition of face, and the facial image after identification is uploaded to cloud
End server carries out matching verifying, when matching authentication failed, issues alarm command to warning device.
It should be noted that the setting of intensity threshold is in laboratory conditions, to control indoor light in the present embodiment,
By being set on the recognition effect for the RGB image for acquiring face.When identifying that error rate is higher on RGB image, at this time
Can substantially be determined as that current light intensity not can guarantee RGB image being capable of discrimination with higher.But face at this time
Profile still can detecte out, that is, still can detecte out the region where face on RGB image, can borrow at this time
It helps 3D point cloud to be identified, improves recognition accuracy.
It should be noted that the present embodiment in the preferable situation of light, directly carries out the knowledge of 2D image using RGB image
Not;When light is poor, Face datection is carried out using RGB image, corresponding face three-dimensional point cloud is then extracted and carries out face knowledge
Not, i.e., 2D image is only to provide detection face effect, and identifies without 2D, carries out recognition of face using 3D point cloud, improves
The stability and accuracy of different light intensity human face identifications.
Specifically, as shown in Figure 1, the luminous intensity computing unit 201 includes luminous intensity conversion subunit and light intensity meter
Operator unit;The luminous intensity conversion subunit is used to the RGB image being converted into HSI image, and extracts the figure in the channel I
Picture;The luminous intensity computation subunit be used to be calculated according to the image in the channel I the average value of the image in the channel I and as
The light intensity value of current environment.
Here gone unlike induction environment light from traditional using photosensitive sensor, this programme by the intensity of image come
The intensity of characterization light indirectly.Because light is weaker, the intensity of the image of camera shooting is just smaller.RGB image can only phenogram
The color of picture, but these information contain strength component, by the way that rgb space is transformed into HSI (H: tone, S: saturation degree, I: strong
Degree), environmental light intensity angle value can be calculated according to the image in the channel I, without reducing cost by other sensors.
Specifically, when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, the RGB image is utilized
Carrying out recognition of face is the 2D recognition of face based on RGB image, as shown in Figure 1, the first face identification unit 203 includes the first instruction
Practice collection subelement, the first extraction subelement and the first face coupling subelement.
The first training set subelement is used to adopt in advance when ambient light intensity is greater than or equal to the intensity threshold of setting
The RGB image and depth image for collecting face constitute the first training dataset, and carry out feature to the RGB image in the first training set
It extracts, obtained first eigenvector is stored to database;
Described first, which extracts subelement, is used for when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, right
The face RGB image acquired in real time carries out feature extraction, obtains feature vector;
First face coupling subelement is for carrying out the first eigenvector stored in this feature vector and database
Match, carries out recognition of face.
It should be noted that when the first face identification unit 203 carries out the 2D recognition of face based on RGB image, specifically
Including two stages, model training stage and cognitive phase:
(1) stage-training: when the luminous intensity of current environment is less than the intensity threshold of setting, acquisition in advance needs to pass in and out electricity
The RGB image and depth image of net capital construction scene employee face, and it is configured to the first training dataset, to the first training dataset
In face 2D image carry out feature extraction (such as PCA Principal Component Analysis), then, by the feature of all trained face 2D images
Vector is stored in database module, carries out recognition of face when passing in and out power grid capital construction scene for subsequent employee;
(2) cognitive phase: when current environment luminous intensity is greater than or equal to the intensity threshold of setting, using same feature
Extracting method carries out characteristic vector pickup to the face 2D image that acquires in real time, and by the feature vector and database module of extraction
In feature vector matched, and matching result is exported, it is opening/closing to control power grid capital construction scene gate according to matching result.
Specifically, when the luminous intensity of current environment is less than the intensity threshold of setting, start infrared light compensating lamp, to working as front ring
The infra-red intensity in border carries out light filling, and carrying out recognition of face using the RGB image and depth image is the people based on 3D point cloud
Face identification, the present embodiment can be arranged infrared ray light compensating lamp in RGB-D camera shooting area and additionally give when light is poor
Light filling is given to guarantee the clarity of the infrared figure of light, and infrared ray is human eye observation less than therefore will not generating to human eye any
Stimulation;Second face identification unit 204 includes the first mapping subelement, the second extraction subelement, the second mapping subelement
With the second face coupling subelement.
First mapping subelement is used to acquire face in advance when the luminous intensity of environment is less than the intensity threshold of setting
RGB image and depth image, and extract the region (u on the RGB image where facei, vi) (i=1,2 ... n), and utilizes and reflect
It penetrates formula the RGB image is mapped on the depth image, obtains three dimensional point cloud to construct the second training dataset;
Specifically, then region of the face on RGB image can be obtained with haar Face datection algorithm can obtain according to mapping equation
Take corresponding face three-dimensional point cloud.
The three dimensional point cloud that second extraction subelement is used to concentrate the second training data carries out feature extraction, and will mention
The second feature vector got is stored to database;
Second mapping subelement is used to utilize mapping equation when the luminous intensity of current environment is less than the intensity threshold of setting
The RGB image and depth image of the face acquired in real time are handled, real-time three-dimensional point cloud data is obtained;
Second face coupling subelement is used for the feature that real-time three-dimensional point cloud data feature is extracted, and will be extracted
Vector is matched with the second feature vector in database, carries out recognition of face, and recognition of face is matched authentication failed
As a result it is sent in warning device, to alarm.
The mapping equation is
Wherein, (u, v) is the coordinate value of corresponding points on the RGB image, IrIt is color camera internal reference, T is RGB camera shooting
The module and carriage transformation matrix p=(x, y, z, 1) of [R t] between head coordinate system and depth camera coordinate systemT, indicate that three-dimensional point is sat
Mark, * indicate product.
It should be noted that carrying out face knowledge using the RGB image and depth image in the second face identification unit 204
When Ji not be based on the recognition of face of 3D point cloud, two stages: model training stage and cognitive phase be specifically included:
(1) training stage: when the luminous intensity of current environment is less than the intensity threshold of setting, acquisition in advance needs to pass in and out electricity
The RGB image and depth image of net capital construction scene employee face, the RGB image is mapped on the depth image, is obtained
Three dimensional point cloud carries out feature extraction to construct the second training dataset, to the face 3D point cloud that the second training data is concentrated
Then the feature vector of all trained face 3D point clouds is stored in database module by (such as VFH viewpoint feature histogram method)
In, recognition of face is carried out when passing in and out power grid capital construction scene for subsequent employee.
(2) cognitive phase: when current environment luminous intensity is less than the intensity threshold of setting, using same feature extraction side
Method carries out characteristic vector pickup to the face 3D point cloud that acquires in real time, and by the spy in the feature vector and database module of extraction
Sign vector is matched, and matching result is exported, and it is opening/closing to control power grid capital construction scene gate according to matching result.
It should be noted that ensure that RGB by the way that RGB image and depth image are carried out registration process in the present embodiment
The consistency of image and depth image corresponding points increases the accuracy that recognition of face is carried out subsequently through three-dimensional point cloud.
As shown in Figure 1, the cloud server 3 includes authority checking module 31, compares interface module 32, query interface mould
Block 33 uploads interface module 34 and block chain storage module 35;
Authority checking module 31 is used for the checking request sent according to the data prediction device 2, to the data
Pretreatment unit carries out authority checking;Authority checking module 31 is to be authorized, verified to the interface of calling data center's program
Management, specifically:
(1) the calling third party (such as: serve end program) for needing to call data center's routine interface needs to award first
Power, when authorization, it is desirable to provide the information such as IP, port numbers, Apply Names, call number of called side;After authorization passes through, return to
The unique interface of called side calls token Token;
(2) called side needs to carry out authority checking first in calling interface, and input calls token and called side to believe substantially
Breath, can be verified.
Interface module 32 is compared to locate in advance for data will to be back to by the verification result of authority checking module authority checking
Manage device 2;
Query interface module 32 is used for after the authority checking module verification passes through, and calls data verification interface to described
The automobile ingress/egress record data for meeting inquiry instruction are verified, and are verified as a result, and verification result is back to the number
Data preprocess device, after authentication failed, data prediction device issues alarm command to warning device;
Upload interface module 33 be used for after authority checking module verification passes through for the data prediction device call with
Automobile ingress/egress record data are uploaded to block chain storage module.
The vehicles while passing record data that data prediction device obtains must be with block chain storage module 35 in cloud server
Verifying is compared in the brief information of storage, and cloud server can be returned by comparing interface module 32 to data prediction device
Return whether this record is verified.Query interface module 34 is used for after the authority checking module 31 is verified, and is called
Data verification interface verifies the automobile ingress/egress record data for meeting inquiry instruction, is verified as a result, and will test
Card result is back to the data prediction device 2, if controlling vehicle pass-through by verifying, if unverified, alarming
Device is alarmed, and has illegal vehicle/personnel to pass in and out engineering site with warning.
As shown in Figure 1, warning device 4 includes linkage matching unit 41 and interlink warning unit 42, the linkage matching is single
Member 41 includes unauthorized disengaging alarm part-unit 411, the alarm part-unit 412 that do not wear a safety helmet, abnormal time disengaging alarm son list
Member 413, non-same people pass in and out alarm part-unit 414;
The matching unit 41 that links is used for the matching relationship based on pre-set alarm parameters and linkage parameter, obtains and institute
The corresponding linkage characteristic parameter of current alerts parameter is stated, current alerts parameter is that the alarm that the data prediction device issues refers to
It enables;
Interlink warning unit 42 is sent in warning device corresponding with linkage characteristic parameter for the characteristic parameter that will link,
To realize that early warning links;
Unauthorized disengaging alarm part-unit 411 is used to judge whether the type of alarm of acquired alert data to be unauthorized
Disengaging alarm, if so, linkage characteristic parameter is sent in corresponding warning device with the first alert levels and carries out linkage report
It is alert, if it is not, then entering the alarm part-unit 412 that do not wear a safety helmet;
The alarm part-unit 412 that do not wear a safety helmet is for judging whether type of alarm is the alarm that do not wear a safety helmet;If so, connection
Dynamic characteristic parameter is sent in corresponding warning device with the second alert levels carries out interlink warning, if it is not, then entering abnormal
Time passes in and out alarm part-unit 413;
Abnormal time disengaging alarm part-unit 413 is used to judge whether type of alarm to be abnormal time disengaging alarm, if so,
The characteristic parameter that then links is sent in corresponding warning device with third alert levels carries out interlink warning, if it is not, then entering
Non- same people passes in and out alarm part-unit 414;
Non- same people passes in and out alarm part-unit 414 and is used to judge whether the type of alarm of acquired alert data to be non-same
One people passes in and out alarm, if so, linkage characteristic parameter is sent in corresponding warning device with the 4th alert levels and is joined
Dynamic alarm, if it is not, then without alarm.
Alarm part-unit 411, the alarm part-unit 412 that do not wear a safety helmet, abnormal time disengaging alarm are passed in and out by unauthorized
Subelement 413, non-same people pass in and out alarm part-unit 414 and are alarmed accordingly in engineering site, realize to each alarm class
Type carries out corresponding early warning linkage, but more than listed above these of type of alarm, expansible other type of alarm of access, example
Such as illegal vehicle is alarmed, and realizes and a variety of type of alarm in engineering site are detected and alarmed.Above first alarm
Rank, the second alarm grade and the corresponding linkage alarm device of third alert levels can be to be single extremely multiple, to be able to achieve prompting
Carry out processing of the related personnel after alarming is preferable mode;In the present embodiment, according to the important journey of type of alarm
Degree, preferably the first alert levels and the second alert levels are the connection that sound-light alarm, large-size screen monitors display alarm and mobile terminal alarm combine
Dynamic type of alarm can be set into close gate inhibition or label of expressing that no through traffic at this time, and also can be set into does not allow to pass in and out work
Other information warnings of journey, third alert levels are the interlink warning mode of sound-light alarm, large-size screen monitors display alarm combination, at this time may be used
It is current after current or warning to be arranged to carry out related personnel or vehicle etc. after abnormal time disengaging is registered, also it can be set into
Abnormal time passes in and out other information warnings of engineering, and the 4th alert levels are that large-size screen monitors show type of alarm, can be set at this time
It is current after directly current or warning, other information warnings that abnormal personnel pass in and out engineering also can be set into.
The alert data of acquisition is matched, alarm linkage mode corresponding with alert data, alarm linkage side are obtained
There are many formulas, can be the single or multiple combinations in sound-light alarm, large-size screen monitors display alarm and mobile terminal alarm etc., realization sound
Sound, display, movement are not limited only to above-mentioned alarm linkage mode, are subject to and are able to achieve preferably alarm.Pass through the above alarm linkage
Mode is alarmed, and when solving construction site and alert event occur, linkage, the processing etc. of corresponding subsequent alarms form report
Alert closed loop process, solves in the complex environment of engineering site, it is difficult to alert data source is determined, so that subsequent report
Alert and processing is affected, and causes the defect of engineering site confusion phenomena.
As shown in figure 4, a kind of engineering site passes in and out management method, for including the following steps in data prediction device
S0001 to S0003:
S0001: the record data for obtaining vehicle or personnel's entry/exit engineering site are pre-processed, and upload to cloud
In server;
S0002: the record matching object command that cloud server returns is obtained, the record matching object command is pre- place
Output result in record data and cloud server after reason after preset record attribute Data Matching;
Data prediction device obtains the record matching object command that cloud server returns, and record matching result is referred to
The object command that it fails to match in order is sent in warning device.
S0003: it sends the record matching object command that it fails to match in warning device, to alarm.
For step S0001: data prediction device to obtain the record data of vehicle or personnel's entry/exit engineering site into
Row pretreatment includes the following steps S0011 to S0015 when the face to personnel carries out identification pretreatment:
S0011: the RGB image and depth image of face are obtained respectively;
It should be noted that can use RGB-D camera carries out man face image acquiring, RGB-D camera is imaged including two classes
Head, one kind are RGB cameras, and for shooting the RGB image of face, another kind of is depth camera, for shooting the depth of face
Spend image.
S0012: detecting the luminous intensity of current environment, obtains the luminous intensity of current environment;
S0013, the luminous intensity of current environment is compared with the Intensity threshold of setting;
S0014, current environment luminous intensity be greater than or equal to setting intensity threshold when, using the RGB image into
Row recognition of face;
S0015, current environment luminous intensity be less than setting intensity threshold when, utilize the RGB image and depth map
As carrying out recognition of face.
Above-mentioned steps S0012: detecting the luminous intensity of current environment, obtains the luminous intensity of current environment, including such as
Lower subdivided step S0112-S0212:
S0112: the RGB image is converted into HSI image, and extracts the image in the channel I;
S0212: according to the image in the channel I, the average value of the image in the channel I and the light intensity as current environment are calculated
Angle value.
Above-mentioned steps S0014: when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, using described
It is the 2D face identification method based on RGB image that RGB image, which carries out recognition of face, includes the following steps S0114-S0314:
S0114, ambient light intensity be greater than or equal to setting intensity threshold when, in advance acquire face RGB image and
Depth image constitutes the first training dataset, and carries out feature extraction, first will obtained to the RGB image in the first training set
Feature vector is stored to database;
S0214, current environment luminous intensity be greater than or equal to setting intensity threshold when, to the face acquired in real time
RGB image carries out feature extraction, obtains feature vector;
S0314, the first eigenvector stored in this feature vector and database is matched, carries out recognition of face.
Above-mentioned steps S0015: when the luminous intensity of current environment is less than the intensity threshold of setting, the RGB image is utilized
Carrying out recognition of face with depth image is the recognition of face based on 3D point cloud, including following subdivided step S0115-S0515:
S0115: when the luminous intensity of environment is less than the intensity threshold of setting, the RGB image and depth of face are acquired in advance
Image;
S0215: the region on the RGB image where face is extracted, and is mapped the RGB image using mapping equation
Onto the depth image, three dimensional point cloud is obtained to construct the second training dataset;
Specifically, region of the face on RGB image, then, root can be specifically obtained using haar Face datection algorithm
According to the available corresponding face three-dimensional point cloud of mapping equation.Mapping equation are as follows:
Wherein, (u, v) is the coordinate value of corresponding points on the RGB image, IrIt is color camera internal reference, T is RGB camera
The module and carriage transformation matrix p=(x, y, z, 1) of [R t] between coordinate system and depth camera coordinate systemT, indicate three-dimensional point coordinate, *
Indicate product.
It should be noted that ensure that RGB by the way that RGB image and depth image are carried out registration process in the present embodiment
The consistency of image and depth image corresponding points increases the accuracy that recognition of face is carried out subsequently through three-dimensional point cloud.
S0315, feature extraction is carried out to the three dimensional point cloud that the second training data is concentrated, and special by extract second
Sign vector is stored to database;
S0415, the RGB image and depth image that acquire in real time are handled using the mapping equation, is obtained in real time
Three dimensional point cloud;
S0515, real-time three-dimensional point cloud data feature is extracted, and will be in feature vector and database that extract
Second feature vector is matched, and recognition of face is carried out.
The step S0003: it is sent in warning device by the record matching object command that it fails to match, to be reported
When alert, including S0013 to S0023:
S0013: the matching relationship based on pre-set alarm parameters and linkage parameter obtains joining with the current alerts
The corresponding linkage characteristic parameter of number, current alerts parameter is the alarm command that the data prediction device issues;
S0023: sending linkage characteristic parameter in warning device corresponding with linkage characteristic parameter, to realize that early warning joins
It is dynamic, as shown in fig. 6, including the following steps S0123 to S0823:
S0123: whether the type of alarm for judging acquired alert data is unauthorized disengaging alarm;
If so, entering step S0223;
If not, then it represents that for warning device corresponding with unauthorized disengaging alarm without alarm, disengaging is normal, enters step
S0323;
S0223: linkage characteristic parameter is sent in corresponding warning device with the first alert levels, to carry out early warning connection
It is dynamic;First alert levels are one or more linkage reports combined in sound-light alarm, large-size screen monitors display alarm and mobile terminal alarm
Police's formula, the first alert levels are preferably the interlink warning mode that sound-light alarm, large-size screen monitors display alarm and mobile terminal alarm combine;
S0323: judge whether type of alarm is the alarm that do not wear a safety helmet;
If so, entering step S0423;
If not, then it represents that corresponding warning device is alarmed without alarm with not wearing a safety helmet, safety cap wearing is normal, into
Enter step S0523;
S0423: linkage characteristic parameter is sent in corresponding warning device with the second alert levels, to carry out early warning connection
It is dynamic;
S0523: judge whether type of alarm is abnormal time disengaging alarm, and the abnormal time is to be not at setting time
Time in range;
If so, entering step S0623;
S0623: linkage characteristic parameter is sent in corresponding warning device with third alert levels, to carry out early warning connection
It is dynamic;
If not, then it represents that for warning device corresponding with abnormal time disengaging alarm without alarm, the disengaging time is normal, into
Enter step S0723.
S0723: whether the type of alarm for judging acquired alert data is that non-same people passes in and out alarm;
If so, entering step S0823;
If it is not, then show to pass in and out the corresponding warning device of alarm without alarm with non-same people, it is normal to pass in and out personnel, complete
It links at engineering entirety early warning.
S0823: linkage characteristic parameter is sent in corresponding warning device with the 4th alert levels, to carry out early warning connection
It is dynamic.
Above step S0123 is into S0823, by type of alarm: unauthorized pass in and out alarm, the alarm that do not wear a safety helmet,
Abnormal time disengaging alarm, non-same people pass in and out alarm etc. and judge and carry out corresponding early warning linkage, realize to each report
Alert type carries out corresponding early warning linkage, but more than listed above these of type of alarm, other alarm classes of expansible access
Type, such as illegal vehicle alarm etc., realize and a variety of type of alarm in engineering site are detected and alarmed.
As shown in figure 5, a kind of engineering site passes in and out management method, in cloud server, comprising:
S1001: the record data after obtaining data prediction, the pretreated record data are pretreatment units pair
Acquired record data carry out pretreated data;
S1002: by pretreated record data and preset record attribute Data Matching, by record matching result
Instruction is issued to pretreatment unit, the instruction that it fails to match in the record matching object command for trigger warning device with into
Row alarm.
Those of ordinary skill in the art will appreciate that: realize that all or part of the steps of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer readable storage medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed;And storage medium above-mentioned includes: ROM, RAM, magnetic disk or light
The various media that can store program code such as disk.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (10)
1. a kind of engineering site in-out management system, which is characterized in that including data acquisition facility, data prediction device, cloud
Hold server and warning device;
Data acquisition facility is used to obtain the record data of vehicle or personnel's entry/exit engineering site, and will be in the record data
Data prediction device is passed to, the record data include license plate image data and driver's image data;
Data prediction device uploads in cloud server for pre-processing to the record data;
The cloud server is used for pretreated record data and preset record attribute Data Matching, and will note
It records matching result and instructs returned data pretreatment unit;
Warning device is used to obtain data prediction device and issues the record data result instruction that it fails to match, to be reported
It is alert.
2. engineering site in-out management system according to claim 1, which is characterized in that the data prediction device packet
It includes for carrying out the pretreated face image processing module of recognition of face to personnel, the face image processing module includes light intensity
Computing unit, luminous intensity comparing unit, the first face identification unit and the second face identification unit are spent, the luminous intensity calculates single
Member includes luminous intensity conversion subunit and luminous intensity computation subunit;
The luminous intensity computing unit obtains the luminous intensity of current environment for detecting to the luminous intensity of current environment;
Luminous intensity comparing unit is for the luminous intensity of current environment to be compared with the Intensity threshold of setting;
First face identification unit is used to utilize data when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting
The RGB image that acquisition device uploads carries out recognition of face, and the facial image after identification is uploaded to cloud server and is matched
Verifying, when matching authentication failed, issues alarm command to warning device;
Second face identification unit is used to obtain when the luminous intensity of current environment is less than the intensity threshold of setting using the data
The RGB image and depth image for taking device to upload carry out recognition of face, and the facial image after identification is uploaded to cloud server
Matching verifying is carried out, when matching authentication failed, issues alarm command to warning device;
The luminous intensity conversion subunit is used to the RGB image being converted into HSI image, and extracts the image in the channel I;
The luminous intensity computation subunit be used to be calculated according to the image in the channel I the average value of the image in the channel I and as
The light intensity value of current environment.
3. engineering site in-out management system according to claim 2, which is characterized in that the first face identification unit includes
First training set subelement, the first extraction subelement and the first face coupling subelement, second face identification unit include
First mapping subelement, second extract subelement, the second mapping subelement and the second face coupling subelement;
The first training set subelement is used to acquire people in advance when ambient light intensity is greater than or equal to the intensity threshold of setting
The RGB image and depth image of face constitute the first training dataset, and carry out feature to the RGB image in the first training set and mention
It takes, obtained first eigenvector is stored to database;
Described first, which extracts subelement, is used for when the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, to real-time
The face RGB image of acquisition carries out feature extraction, obtains feature vector;
First face coupling subelement is used to match the first eigenvector stored in this feature vector and database, into
Row recognition of face, and send the result of recognition of face matching authentication failed in warning device, to alarm;
First mapping subelement is used to acquire the RGB figure of face in advance when the luminous intensity of environment is less than the intensity threshold of setting
Picture and depth image, and the region on the RGB image where face is extracted, and the RGB image is mapped using mapping equation
Onto the depth image, three dimensional point cloud is obtained to construct the second training dataset;
The three dimensional point cloud that second extraction subelement is used to concentrate the second training data carries out feature extraction, and will extract
Second feature vector store to database;
Second mapping subelement is used for when the luminous intensity of current environment is less than the intensity threshold of setting, using mapping equation to reality
When the RGB image of face that acquires and depth image handled, obtain real-time three-dimensional point cloud data;
Second face coupling subelement is used for the feature vector that real-time three-dimensional point cloud data feature is extracted, and will be extracted
It is matched with the second feature vector in database, carries out recognition of face, and recognition of face is matched to the result of authentication failed
It is sent in warning device, to alarm;
The mapping equation are as follows:
Wherein, (u, v) is the coordinate value of corresponding points on the RGB image, IrIt is color camera internal reference, T is RGB camera coordinate
The module and carriage transformation matrix p=(x, y, z, l) of [R t] between system and depth camera coordinate systemT, indicate three-dimensional point coordinate, * is indicated
Product.
4. engineering site in-out management system according to claim 1, which is characterized in that the cloud server includes awarding
Authentication module is weighed, interface module is compared, query interface module, uploads interface module and block chain storage module;
Authority checking module is used for the checking request sent according to the data prediction device, to fill to the data prediction
Set carry out authority checking;
Interface module is compared for data prediction device will to be back to by the verification result of authority checking module authority checking;
Query interface module be used for after the authority checking module verification passes through, call data verification interface to it is described meet look into
The automobile ingress/egress record data for asking instruction are verified, and are verified as a result, and verification result is back to the server
End;
Interface module is uploaded to be used to after authority checking module verification passes through calling for the data prediction device with will be described
Automobile ingress/egress record data are uploaded to block chain storage module.
5. engineering site in-out management system according to claim 1, which is characterized in that the data acquisition facility includes
Image collecting device and the image conversion apparatus being connect with image collecting device output end;
Described image acquisition device includes to the first camera identified by gate entry/exit vehicle, for passing through lock
The driver personnel of machine entry/exit vehicle carry out the second camera of recognition of face and for the non-through vehicles for crossing gate entry/exit
Or the third camera that is identified of personnel, the output end of the first camera, second camera and third camera connect
It is connected to the input terminal of image conversion apparatus;
Image conversion apparatus includes video decoding circuit and is filtered the first filter circuit of noise reduction, the first filtered electrical to image
The output end on road and the input terminal of video decoding circuit connect;
Video decoding circuit includes video decoding chip, crystal oscillating circuit and the second filter circuit, the output end connection of crystal oscillating circuit
To the input terminal of video decoding chip, the input terminal of the second filter circuit is connected to the output end of external power supply, the second filtered electrical
The output end on road is connected to the analogue ground of frequency decoding chip;
First filter circuit includes capacitor and two resistance, wherein a resistance one end is connected to the output end, another of interface P1
End is connected respectively to capacitance cathode end and another resistance one end, the other end ground connection of another resistance, and capacitor negative pole end is connected to institute
It states on the input pin of video decoding chip;
Second filter circuit includes at least one filter capacitor, is connected in parallel to each other between filter capacitor, the positive terminal after parallel connection
It is connected to external power supply end, the negative pole end after parallel connection is connected to the analogue ground pin of the video decoding chip.
6. engineering site in-out management system according to claim 1, which is characterized in that warning device includes linkage matching
Unit and interlink warning unit, the linkage matching unit includes unauthorized disengaging alarm part-unit, do not wear a safety helmet son of alarming
Unit, abnormal time disengaging alarm part-unit, non-same people pass in and out alarm part-unit;
The matching unit that links is used for the matching relationship based on pre-set alarm parameters and linkage parameter, obtains and current alerts
The corresponding linkage characteristic parameter of parameter, current alerts parameter are the alarm commands that the data prediction device issues;
Interlink warning unit is sent in warning device corresponding with linkage characteristic parameter for the characteristic parameter that will link, to realize
Early warning linkage;
Unauthorized disengaging alarm part-unit is used to judge whether the type of alarm of acquired alert data to be unauthorized disengaging report
It is alert, if so, linkage characteristic parameter is sent in corresponding warning device with the first alert levels and carries out interlink warning, if
It is no, then enter the alarm part-unit that do not wear a safety helmet;
The alarm part-unit that do not wear a safety helmet is for judging whether type of alarm is the alarm that do not wear a safety helmet;If so, linkage feature
Parameter is sent in corresponding warning device with the second alert levels and carries out interlink warning, if it is not, then enter abnormal time into
Alarm part-unit out;
Abnormal time disengaging alarm part-unit is for judging whether type of alarm is abnormal time disengaging alarm, if so, linkage
Characteristic parameter is sent in corresponding warning device with third alert levels and carries out interlink warning, if it is not, then entering non-same
People passes in and out alarm part-unit;
Non- same people pass in and out alarm part-unit be used to judge the type of alarm of acquired alert data whether be non-same people into
It alarms out, if so, linkage characteristic parameter is sent in corresponding warning device with the 4th alert levels and carries out interlink warning,
If it is not, then without alarm.
7. a kind of engineering site passes in and out management method, which is characterized in that in data prediction device, comprising:
The record data for obtaining vehicle or personnel's entry/exit engineering site are pre-processed, and are uploaded in cloud server;
The record matching object command that cloud server returns is obtained, the record matching object command is pretreated record
Output result in data and cloud server after preset record attribute Data Matching;
It sends the record matching object command that it fails to match in warning device, to alarm.
8. engineering site according to claim 7 passes in and out management method, which is characterized in that acquisition vehicle or personnel
When the record data of entry/exit engineering site are pre-processed, when carrying out identification matching to face, comprising:
The RGB image and depth image of face are obtained respectively;
The luminous intensity of current environment is detected, the luminous intensity of current environment is obtained;
The luminous intensity of current environment is compared with the Intensity threshold of setting;
When the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, recognition of face is carried out using the RGB image,
Facial image after identification is uploaded into cloud server and carries out matching verifying, when matching authentication failed, under warning device
Send out alarm command;
When the luminous intensity of current environment is less than the intensity threshold of setting, face is carried out using the RGB image and depth image
Identification, uploads to cloud server for the facial image after identification and carries out matching verifying, when matching authentication failed, fills to alarm
It sets and issues alarm command.
9. a kind of engineering site passes in and out management method, in cloud server, comprising:
Record data after obtaining data prediction, the pretreated record data are pretreatment units to acquired note
It records data and carries out pretreated data;
By pretreated record data and preset record attribute Data Matching, record matching object command is issued to
Pretreatment unit, the instruction that it fails to match in the record matching object command is for triggering warning device to alarm;
10. engineering site according to claim 9 passes in and out management method, which is characterized in that acquisition vehicle or personnel
When the record data of entry/exit engineering site are pre-processed, when carrying out identification matching to face, comprising:
The RGB image and depth image of face are obtained respectively;
The luminous intensity of current environment is detected, the luminous intensity of current environment is obtained;
The luminous intensity of current environment is compared with the Intensity threshold of setting;
When the luminous intensity of current environment is greater than or equal to the intensity threshold of setting, recognition of face is carried out using the RGB image,
Facial image after identification is uploaded into cloud server and carries out matching verifying, when matching authentication failed, under warning device
Send out alarm command;
When the luminous intensity of current environment is less than the intensity threshold of setting, face is carried out using the RGB image and depth image
Identification, uploads to cloud server for the facial image after identification and carries out matching verifying, when matching authentication failed, fills to alarm
It sets and issues alarm command.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910666189.0A CN110458985A (en) | 2019-07-19 | 2019-07-19 | A kind of engineering site in-out management system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910666189.0A CN110458985A (en) | 2019-07-19 | 2019-07-19 | A kind of engineering site in-out management system and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110458985A true CN110458985A (en) | 2019-11-15 |
Family
ID=68483079
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910666189.0A Pending CN110458985A (en) | 2019-07-19 | 2019-07-19 | A kind of engineering site in-out management system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110458985A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111554006A (en) * | 2020-04-13 | 2020-08-18 | 绍兴埃瓦科技有限公司 | Intelligent lock and intelligent unlocking method |
CN111737674A (en) * | 2020-08-03 | 2020-10-02 | 德能森智能科技(成都)有限公司 | Wisdom garden system based on cloud platform |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104320881A (en) * | 2014-10-28 | 2015-01-28 | 江苏天语雅思医疗设备有限公司 | Intelligent dimming controller in LED shadowless lamp lighting system |
CN107845174A (en) * | 2017-11-23 | 2018-03-27 | 郑州云海信息技术有限公司 | A kind of kindergarten's gate control system and method based on cloud computing |
CN207264587U (en) * | 2017-08-04 | 2018-04-20 | 国网浙江省电力公司湖州供电公司 | Engineering site personnel's Vehicular system |
CN207780803U (en) * | 2017-12-18 | 2018-08-28 | 信利光电股份有限公司 | A kind of face identification device |
KR101975962B1 (en) * | 2018-08-06 | 2019-05-08 | 이재일 | Management method for unmanned accommodation |
-
2019
- 2019-07-19 CN CN201910666189.0A patent/CN110458985A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104320881A (en) * | 2014-10-28 | 2015-01-28 | 江苏天语雅思医疗设备有限公司 | Intelligent dimming controller in LED shadowless lamp lighting system |
CN207264587U (en) * | 2017-08-04 | 2018-04-20 | 国网浙江省电力公司湖州供电公司 | Engineering site personnel's Vehicular system |
CN107845174A (en) * | 2017-11-23 | 2018-03-27 | 郑州云海信息技术有限公司 | A kind of kindergarten's gate control system and method based on cloud computing |
CN207780803U (en) * | 2017-12-18 | 2018-08-28 | 信利光电股份有限公司 | A kind of face identification device |
KR101975962B1 (en) * | 2018-08-06 | 2019-05-08 | 이재일 | Management method for unmanned accommodation |
Non-Patent Citations (1)
Title |
---|
邵静: ""基于立体匹配的低纹理图像三维点云重建算法研究"", 《信息科技辑》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111554006A (en) * | 2020-04-13 | 2020-08-18 | 绍兴埃瓦科技有限公司 | Intelligent lock and intelligent unlocking method |
CN111554006B (en) * | 2020-04-13 | 2022-03-18 | 绍兴埃瓦科技有限公司 | Intelligent lock and intelligent unlocking method |
CN111737674A (en) * | 2020-08-03 | 2020-10-02 | 德能森智能科技(成都)有限公司 | Wisdom garden system based on cloud platform |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109214276A (en) | A kind of system and method for the target person track following based on face recognition technology | |
CN107730898A (en) | Parking lot illegal vehicle recognition methods and system | |
CN107705574A (en) | A kind of precisely full-automatic capturing system of quick road violation parking | |
CN105913037A (en) | Face identification and radio frequency identification based monitoring and tracking system | |
CN106709468A (en) | City region surveillance system and device | |
CN104240349A (en) | Method for quickly confirming real name identity in important place and human image and identity comparison safety inspection system | |
CN202838681U (en) | Security intelligent checkpoint system | |
CN104867217A (en) | Community monitoring and management system | |
CN107220633A (en) | A kind of intelligent mobile enforcement system and method | |
CN110213501A (en) | A kind of grasp shoot method, device, electronic equipment and storage medium | |
CN103186982A (en) | Intelligent access system for public security | |
CN204087314U (en) | Portrait identity comparison safe examination system | |
CN204155450U (en) | High definition bayonet apparatus on a kind of road | |
CN105844245A (en) | Fake face detecting method and system for realizing same | |
CN110070043A (en) | It is a kind of that training management system and method is driven based on recognition of face | |
CN110458985A (en) | A kind of engineering site in-out management system and method | |
CN107610393A (en) | A kind of intelligent office monitoring system | |
CN106791655A (en) | A kind of method for processing video frequency and device | |
CN112464030B (en) | Suspicious person determination method and suspicious person determination device | |
CN106548113A (en) | Image-recognizing method and system | |
CN105528626A (en) | RFID (Radio Frequency Identification) reader and camera integrated machine and application thereof | |
CN105608464A (en) | Face recognition mobile terminal solving system and solving method | |
CN111798667A (en) | Monitoring method, monitoring server, monitoring system and storage medium of non-motor vehicle | |
JP2020518165A (en) | Platform for managing and validating content such as video images, pictures, etc. generated by different devices | |
CN108805184B (en) | Image recognition method and system for fixed space and vehicle |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191115 |
|
RJ01 | Rejection of invention patent application after publication |