CN109300200A - A kind of artificial intelligence visitor management system with face recognition - Google Patents
A kind of artificial intelligence visitor management system with face recognition Download PDFInfo
- Publication number
- CN109300200A CN109300200A CN201810334327.0A CN201810334327A CN109300200A CN 109300200 A CN109300200 A CN 109300200A CN 201810334327 A CN201810334327 A CN 201810334327A CN 109300200 A CN109300200 A CN 109300200A
- Authority
- CN
- China
- Prior art keywords
- visitor
- reservation
- eye
- point
- face
- 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
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 17
- 230000005764 inhibitory process Effects 0.000 claims abstract description 8
- 238000012544 monitoring process Methods 0.000 claims abstract description 8
- 210000004279 orbit Anatomy 0.000 claims description 43
- 238000000034 method Methods 0.000 claims description 15
- 230000001815 facial effect Effects 0.000 claims description 13
- 238000010606 normalization Methods 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 11
- 210000004709 eyebrow Anatomy 0.000 claims description 8
- 239000013589 supplement Substances 0.000 claims description 8
- 238000003801 milling Methods 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 5
- 210000001747 pupil Anatomy 0.000 claims description 4
- 210000004209 hair Anatomy 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 2
- 230000008030 elimination Effects 0.000 claims 1
- 238000003379 elimination reaction Methods 0.000 claims 1
- 238000012800 visualization Methods 0.000 claims 1
- 238000012937 correction Methods 0.000 description 10
- 238000013528 artificial neural network Methods 0.000 description 5
- 210000000887 face Anatomy 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 239000013598 vector Substances 0.000 description 3
- 108091026890 Coding region Proteins 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000012790 confirmation Methods 0.000 description 2
- 238000013079 data visualisation Methods 0.000 description 2
- 230000009977 dual effect Effects 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000005693 optoelectronics Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C9/00—Individual registration on entry or exit
- G07C9/20—Individual registration on entry or exit involving the use of a pass
- G07C9/22—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder
- G07C9/25—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder using biometric data, e.g. fingerprints, iris scans or voice recognition
- G07C9/257—Individual registration on entry or exit involving the use of a pass in combination with an identity check of the pass holder using biometric data, e.g. fingerprints, iris scans or voice recognition electronically
Abstract
A kind of artificial intelligence visitor management system with face recognition, including intelligent terminal, external network server, intranet server, access control equipment, vehicle management equipment and payment devices, visitor accesses external network server by intelligent terminal, access control equipment includes monitoring camera, fingerprint acquisition device, bio-identification access controller and electronic lock, the real-time visitor's face-image of monitoring camera dynamic acquisition, fingerprint collecting equipment acquires real-time visitor's fingerprint, bio-identification access controller can identify real-time visitor's face-image and real-time visitor's fingerprint, and carry out matching comparison, after if the two all compares successfully, open electronic lock, visitor is allowed to pass through gate inhibition, if the two compares repeatedly failure, it can starting alarm in real time.
Description
Technical field
The invention belongs to visitor management system field, in particular to a kind of artificial intelligence visitor pipe with face recognition
Reason system.
Background technique
By the way that investigating on the spot, current visitor management system is had the following problems: visitor to certain department is lined up and accesses
Registration, queuing take a long time;When greffier carries out verification ID card information, visitor's typing, it need to first make a phone call to carry out with interviewed people
Confirmation, formality are cumbersome;Visitor takes visitor and temporarily demonstrate,proves, and reaches at lock, and whether hall entrance guard examination of document of being responsible for swiping the card may be used
With not verified to visitor's identity, there are security risks;Current guest system takes a long time for visitor, hand
Continue cumbersome, easily delays visitor and invite the time, while leaving undesirable impression to visitor, reduce department's work efficiency, cause to take
Business low SI;Current vehicle management remains unchanged using the papery pass of backwardness and the two ways of phone confirmation, safety
Lower and timeliness is poor;Eatery Consumption still uses traditional IC card, inconvenient for use and easy to damage, loss.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of artificial intelligence visitor management system with face recognition,
The image of the living body faces image and database purchase that acquire in real time can be effectively identified by face recognition algorithm
Between difference.
To achieve the goals above, the technical solution of the present invention is as follows: a kind of artificial intelligence visitor pipe with face recognition
Reason system, including intelligent terminal, external network server, intranet server, access control equipment, vehicle management equipment and payment devices,
Visitor accesses external network server by intelligent terminal, carries out reservation application, examining department is by network server to the pre- of visitor
It is about examined, and approval results is fed back into external network server and intranet server respectively, external network server is tied according to examination & approval
Fruit generates reservation feedback and is sent to intelligent terminal, and intranet server is set with access control equipment, vehicle management equipment and payment respectively
Standby connection, if reservation is fed back through, the visitor for allowing to hold reservation feedback is entered Office Area, vehicle pipe by access control equipment
The vehicle that reason equipment allows to hold reservation feedback enters parking lot, and payment devices will allow to hold the visitor of reservation feedback
Carry out payment activity;Access control equipment includes monitoring camera, fingerprint acquisition device, bio-identification access controller and electronics
Lock, the real-time visitor's face-image of monitoring camera dynamic acquisition, fingerprint collecting equipment acquire real-time visitor's fingerprint, bio-identification
Access controller can identify real-time visitor's face-image and real-time visitor's fingerprint, and carry out matching comparison, if two
After person compares successfully, electronic lock is opened, visitor is allowed starting to report in real time by gate inhibition if the two comparison repeatedly failure
It is alert.
Beneficial effects of the present invention:
1) visitor is approximately a kind of managing caller scheme Internet-based in advance, using the reservation system, visitor it is in office where
Point can be reserved to visiting without falling into a long wait.It improves gate inhibition's environment, simplifies registering flow path, saves
Access time
2) after the guest login system, according to oneself working demand, the reservation situation of visiting department required for inquiring, choosing
Behind the door, visitor can check the recent reservation situation of the department, to select the suitable time to be reserved at middle part.
3) section is added in the rights management of backstage to exempt from reservation function and prevented external office clerks from frequently reserving
Puzzlement;
4) staff for exempting from reservation for section in entrance guard management can open multiple right of access, and then increase
Working efficiency also improves guarantee in secure context, accomplishes have mark that can follow;
5) the visiting department for exempting from the external staff of reservation and visiting number can be done phase by department work staff
The record and statistics answered, facilitate the result queries in later period;
6) the dual identification of face, fingerprint, practicability is high, safe and reliable, and system uses network information encrypted transmission, supports
It is remotely controlled and is managed, can be applied to gate inhibition's security control of key area,
6) by face recognition algorithm, the living body faces image and database purchase acquired in real time can be effectively identified
Image between difference, thus effectively identify user, prevent from cheating gate inhibition by simple non-living body photo;
7) vehicle management is a management system based on identification technology exploitation.It, can be in vehicle information management module
The vehicle new information in database is inquired by data visualization interface, is increased newly, modification, the operation such as deletion improves work
Make efficiency and safety;
8) it uses artificial neural network recognition methods to identify license plate, therefore discrimination is higher, anti-interference
It is good;
9) it is paid using new-type Encryption Algorithm encrypted transaction message and using two dimensional code, ensure that transaction
Under the premise of safety, the convenient of transaction is realized, user experience is improved.
Detailed description of the invention
Fig. 1 is system pie graph of the invention;
Fig. 2 is that the matching of face-image of the invention compares flow chart;
Fig. 3 is the image recognition flow chart of license plate figure of the invention;
Fig. 4 is two dimensional code payment flow figure of the invention;
Specific embodiment
The present invention will be further described below with reference to the accompanying drawings and embodiments.
The embodiment of the present invention is with reference to shown in Fig. 1-4.
A kind of artificial intelligence visitor management system with face recognition, including intelligent terminal, external network server, Intranet
Server, access control equipment, vehicle management equipment and payment devices, visitor access external network server by intelligent terminal, carry out
Reservation application, examining department examines the reservation of visitor by network server, and approval results are fed back to outside respectively
Network server and intranet server, external network server generate reservation feedback according to approval results and are sent to intelligent terminal, Intranet clothes
Business device is connect with access control equipment, vehicle management equipment and payment devices respectively, if reservation is fed back through, access control equipment will
The visitor for allowing to hold reservation feedback enters Office Area, and the vehicle that vehicle management equipment allows to hold reservation feedback enters
The visitor for allowing to hold reservation feedback is carried out payment activity by parking lot, payment devices;Access control equipment includes monitoring camera
Head, fingerprint acquisition device, bio-identification access controller and electronic lock, the real-time visitor's face of monitoring camera dynamic acquisition
Image, fingerprint collecting equipment acquire real-time visitor's fingerprint, bio-identification access controller can to real-time visitor's face-image and
Real-time visitor's fingerprint identifies, and carries out matching comparison, if after the two all compares successfully, opening electronic lock, allows visitor
It, can starting alarm in real time if the two compares repeatedly failure by gate inhibition.
Wherein, the matching comparison process of face-image is as follows:
Step 1, size normalizes, specifically:
Step 1.1,27 facial fiducial points are obtained, be respectively as follows: right pupil, left pupil, nose, the right corners of the mouth, the left corners of the mouth,
Endpoint in right eyebrow the outer end point, right eyebrow, right eye socket of the eye the outer end point, right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point, endpoint, a left side in right eye socket of the eye
Endpoint in eyebrow, left eyebrow the outer end point, endpoint, left eye socket of the eye upper extreme point, left eye socket of the eye lower extreme point, left eye socket of the eye the outer end point, the nasion in left eye socket of the eye
Portion's right endpoint, nasion portion left end point, wing of nose right part, wing of nose left part, wing of nose right endpoint, wing of nose left end point, upper lip upper extreme point,
Upper lip lower extreme point, lower lip upper extreme point, lower lip lower extreme point;
Said reference point is by the extracting section of face most feature, to improve the precision of identification;
Step 1.2, it is determined according to endpoint in right eye socket of the eye the outer end point, right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point, right eye socket of the eye
Right eye central point determines in left eye according to endpoint, left eye socket of the eye upper extreme point, left eye socket of the eye lower extreme point, left eye socket of the eye the outer end point in left eye socket of the eye
Heart point finds the center of face point of right eye central point and left eye central point, using this center of face point as image origin, rotation
Face-image, the line for being linked to be right eye central point and left eye central point is in horizontal position, to rectify face-image;
Step 1.3, facial ratio characteristic is combined according to facial fiducial point, facial major part is cut into, thereafter will
Face-image after cutting passes through the image that scale transformation is uniform sizes size, eliminates such as the hair except facial parts, carries on the back
The redundancies such as scape;
Wherein, redundancy includes hair, background;
Redundancy is deleted, facial characteristics is more highlighted.
Step 1.4, normalization eye socket distance is than variance,
Step 1.4.1 calculates the distance d of right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point;
Step 1.4.2 calculates center of face point and the left and right endpoint midpoint distance l of the wing of nose;
Step 1.4.3 calculates eye socket distance ratio r=d/l, ratio characteristic of the eye socket distance than having reacted face;
Step 1.4.4, according to eye socket distance than as origin uniform zoom and being cut out using the left and right endpoint midpoint of the wing of nose
Cut face-image;
Step 2, gray scale normalization,
Step 2.1, gray scale adjusts,
If size is the grayscale image of W × H, gray matrix function is I (i, j), and I (i, j) represents each pixel
Intensity, gray matrix function adjusted be S (i, j) then,
S (i, j)=c × cos θ × log (1+I (i, j)),
Wherein, c is constant, and θ is shooting angle,
Step 2.2, matrix function is constructed,
Normalization matrix function is G (i, j), then,
Wherein μ is that gray average is before normalizing, and normalization front difference is σ, then, gray average after normalization
For μ0, normalization variance is σ0;
By above-mentioned gray scale normalization step, make gray level image more standard, is conducive to improve the accuracy compared.
Step 3, histogram is constructed,
Step 3.1, the space λ is constructed, the gradient magnitude T (i, j) and deflection A (i, j) of each pixel, enhancing wheel are calculated
Wide information,
Wherein, Tx(i, j) and Ty(i, j) respectively represents the gradient magnitude on both horizontally and vertically,
Step 3.2, centered on each facial fiducial point, the pixel square region that 27 sizes are 100 × 100 is constructed,
Construct the histogram component function h of every rowj(i, j)=A (i, j) × T (i, j), i=1,2......100, j=1,
2......100, histogram component is hj=[hj(1,j),hj(2,j)......hj(i,j),......hj(100, j)], building is every
The histogram component H of a square regionk=[h1,h2...hj...h100], k=1,2......27;
Above-mentioned histogram covers much information, and constitutes single vector-quantities, and with calculating, improve arithmetic speed.
Step 4, similarity is calculated,
By step 1-3 by real-time visitor's face-image and reservation upload mug shot handled to obtain it is respective straight
Square component, respectivelyWithBy comparing two histogram components, so that similarity between the two is calculated, it is specific to calculate
It constitutes as follows:
Successful match is thought if D is less than 0.3, otherwise it fails to match.
By above-mentioned face recognition algorithm, the living body faces image acquired in real time and data inventory can be effectively identified
Difference between the image of storage prevents from cheating gate inhibition by simple non-living body photo to effectively identify user.
The system simplifies registering flow path, has saved access time, and working efficiency improves.
Wherein, the detailed process of the reservation application are as follows: visitor logs in the reservation in external network server by intelligent terminal
System, according to working demand, the reservation situation of visiting department required for inquiring, after choosing visiting department, visitor checks the visiting
The recent reservation situation of department, to select the suitable time to be reserved, visitor fill in reservation information need to be by name, electricity
Words number, identification card number, visiting department, visiting time fill in strictly according to the facts and mug shot and fingerprint are uploaded to reservation system, also
Same administrative staff or information of vehicles can be filled according to demand.
Wherein, information of vehicles includes license plate number, vehicle and color;
Wherein, the detailed process of the reservation feedback is generated are as follows: for visitor after reservation application, examining department passes through reservation
System receives access application immediately, waits examining department's examination & approval, and reservation system is anti-by the approval results generation reservation of examining department
Feedback, the intelligent terminal of visitor is sent to by wechat or SMS platform;
Reservation feedback include the name of visitor, telephone number, identification card number, visiting department, visiting time, mug shot,
Fingerprint, information of vehicles, subscription state, permission,
Feedback content is abundant, is convenient for subsequent data processing and statistical analysis.
Wherein, reservation applies for that the pending time is 30 minutes, if not processed in 30 minutes, reservation system will be according to reservation
Situation is given an written reply automatically and is sent to by wechat or SMS platform the intelligent terminal of visitor to application;
It prevents from declining without user experience caused by answer for a long time.
Wherein, reservation feedback is by by the database of automatic input intranet server, and related personnel and visitor can be with
Pass through platform or wechat query-reservation application at any time;
Using kinds of platform, reserve more flexible and convenient.
Wherein, reservation system includes backstage authority management module, and examining department combines reservation anti-by authority management module
Feedback to visitor carry out priority assignation, allow external staff to exempt to reserve within a certain period of time, operate access control equipment in for
Exempt from the open multiple right of access of staff in the period of reservation, and the visiting portion of the external staff of reservation will be exempted from
Door and visiting number are done in corresponding record and the database of statistics typing intranet server.
The mode for exempting from reservation increases working efficiency and also improves guarantee in secure context.
Wherein, vehicle management equipment includes that Car license recognition camera, vehicle room entry/exit management device and channel milling train, vehicle go out
Enter manager to compare the license board information in the information of vehicles of Car license recognition cameras capture and the database of intranet server
It is right, after comparing successfully, by license plate number, type of vehicle, vehicle color, entry time, mode of operation, alarm condition and vehicle
Board figure is uploaded in the database of intranet server, and vehicle room entry/exit management device control channel milling train is let pass, and is repeatedly lost if compared
Lose, vehicle room entry/exit management device then can automatic upload information to administrative staff.
Information of vehicles is comprehensive, convenient for the management of administrative staff.
Wherein, channel milling train includes restrictor bar, collet, fork arm, cabinet, chassis lid, motor, retarder, belt wheel, gear, company
Bar, distant bar, main shaft, balancing spring, optoelectronic switch, control box, on the cabinet of Car license recognition camera installation passage milling train, acquisition
It is substantially flush when image photographic with the height of vehicle license plate.
Milling train and camera integrated setting are realized, and horizontal shooting, angle are more acurrate.
Wherein, Car license recognition cameras capture information of vehicles and in the database of intranet server license board information carry out
Compare include Car license recognition cameras capture license plate figure after image recognition with the license plate number in the database of intranet server
Code compares, type of vehicle compares and vehicle color compares,
Wherein, the detailed process of the image recognition of license plate figure are as follows:
Step 1, gradation conversion;
Step 2, contrast equalizes, if g (i, j), (i=1,2 ..., M;J=1,2 ..., N) for after gradation conversion
Image, wherein M, N are respectively the height and width on image pixel dimensions, and the grey scale change range of image is [0,255], specifically
Are as follows:
Step 2.1, according to original image [f (i, j)]M×NThe h of 256 dimension of buildingf(t), t=0,1,2 ..., 255 vectors;
Step 2.2 seeks the intensity profile Probability p of original imagef(t) vector,
Wherein, NfFor total number of pixels of image;
Step 2.3, the cumulative distribution Probability p of each gray value of image is calculateda(k), then,
Wherein, pa(0)=0;
Step 2.4, histogram equalization calculates, the pixel value g (i, j) of image after being handled, then, and g (i, j)=
255*pa(k)。
Therefore the greyscale transformation value of each pixel after equilibrium can be acquired according to the statistic of original image, the figure after equalization
Image contrast is strengthened, and avoids generating interference to license plate part, the license plate of uneven illumination is made to become relatively clear.
Step 3, median filtering is carried out to gray level image, carries out edge detection using Canny operator and gradient operator, makes
Boundary point is eliminated with etching operation, highlights license plate area;
Median filtering can be effectively protected image border, and can remove noise;Canny operator is dual threshold side
Method is not easily susceptible to noise jamming, can detect real weak edge;Gradient operator can be detected along specific direction, gradient operator
Operand and data volume can be considerably reduced with the combination of both Canny operator, effectively removes incoherent parts of images.
Step 4, License Plate;
Step 4.1, building structure matrix carries out closed operation to image, and license plate area is filled into a connected region, is disappeared
Except small region, it is connected to license plate part, indicates preselected area out;
Step 4.2, white is set by preselected area, pixel value 0, other regions are set as black, pixel value 1;It is right
The closed white block of each of image is labeled, specifically:
Image is scanned, encountering first pixel value is 1 to be labeled as LAB, then to the object of its 8 neighborhood into
Row scanning, if this eight pixel values are 1, these pixels all in a region, are labeled as LAB, then proceed to sweep
It retouches, indicates that this region has marked when scanning to pixel value is 0 pixel and finish, be 1 when pixel value is arrived in scanning again
Point when, LAB=LAB+1 is successively scanned whole image;
Step 4.3, the length-width ratio of tab area is calculated, specifically:
After having marked all areas, the height and width in each region are calculated, by the extreme value up and down in each region
Point is respectively labeled as Wmax, Wmin, Hmax, Hmin, then width W=Wmax-Wmin, height H=Hmax-Hmin;
Step 4.4, region is screened, specifically: use the ratio of width to height W/H as constraint condition, by the ratio of width to height W/H in [2,5]
Region in addition all deletes, if the satisfactory region the ratio of width to height W/H have it is multiple, according to absolute width value W and height
Value H, is further reduced interference region;The area for calculating remaining area, the number of pixel in every piece of region is added up and is averaged
Value, selects and removes the connected region less than 500, to obtain license plate area;
Step 4.5, by obtaining the specific location of license plate to the positioning of rectangular area;
Step 5, the inclination angle that rectangular area is detected using Hough transform, further according to the tilt angle that detected into
Row interpolation rotation, keeps the character in rectangular area horizontally arranged;
Camera is adjusted to the position with license plate holding level as far as possible, but does not ensure that license plate is hung with inclination angle, such as
Fruit, which is not corrected, to make a big impact to next step Character segmentation, and Hough transform have to the extraction of straight line it is very strong anti-dry
Disturb ability.
Step 6, character recognition,
Step 6.1, Character segmentation;
Step 6.2, character normalization, specifically: it is obtained when up-and-down boundary and Character segmentation when being extracted according to character
Right boundary, calculate the height and width of each character, the result of calculating counted, using its maximum value as normalizing
The template size of change is tested by calculating, character size is all unified for 50 × 25 pixel size;
Step 6.3, character extracts, specifically: thick meshed feature extraction method is used, character is divided into 10 × 5 nets
Lattice count black picture element number in grid;
Step 6.4, neural network identifies, detailed process are as follows:
Step 6.4.1, designs three BP neural networks to identify the character of license plate, respectively differentiation Chinese character or number and
The BP network of the BP network of English alphabet, the BP network of Chinese character and number and English alphabet, character feature composition characteristic is raw
At feature samples;
Feature samples are input to BP neural network and learnt, establish identification model by step 6.4.2;
Step 6.4.3 identifies characters on license plate using established model;
Step 6.4.4 exports recognition result.
Due to using artificial neural network recognition methods, discrimination is higher, and anti-interference is also relatively good.
Wherein, administrative staff pass through database of the data visualization interface to intranet server of vehicle room entry/exit management device
In information of vehicles the operation such as inquired, increased newly, being modified, being deleted.
Wherein, payment management includes code reader and top-up machines, and top-up machines are according to typing visitor name, telephone number, identity
Card number generates unique visitor's two dimension payment code, is used for identification, visitor is supplemented with money using the payment code by top-up machines
And paid,
Wherein, two dimension payment code includes coding region and function pattern,
Coding region is directly related to information expression, including coded data, release format and Error Correction of Coding;
Function pattern provides auxiliary information, the convenient identification to two-dimension code image, including positioning pattern, separator, position
Detect figure, correction graph;
Wherein, two dimension payment code generating process includes:
Step 1, telephone number, identification card number, payment cipher, payment information are converted into 128 initial data M;
Step 2, data encryption, specifically:
Step 2.1, key is generated,
Select three Big prime a1、a2、a3, modulus n is calculated,
N=a1×a2×a3,
Obtain Euler's function
Select random integers k1, so thatAnd
According to a1、a2、a3Size relation calculates X, GCD (X, n)=1, and n-MAX (a1,a2,a3)≤X≤n,
Obtained X will be solved to substitute into following formula, solution obtains k2;
k2=k1 -1MOD (X),
Then, (k1, X) and it is public key, (k2, X) and it is private key;
Step 2.2, encryption information,
Use public key (k1, X) initial data is encrypted,
C=Mk1MOD (X),
C is encrypted ciphertext;
Wherein, three Big prime a1、a2、a3Binary number digit be greater than 100, the digit of the binary number of modulus n
Greater than 200, three Big prime a1、a2、a3Mutual absolute value of the difference is greater than 50;
The generation of prime number has vital effect to the safety of algorithm, and condition more than satisfaction just can guarantee encryption
Algorithm can resistance factor decomposes in reasonable time range attack.
Step 3, two dimensional code generates,
Step 3.1, data are analyzed, and are analyzed the ciphertext C to be encoded, are determined the type of data, determine data encoding
Mode and error-correction level appropriate determine symbol version to complete the high efficient coding of data;
Step 3.2, data encoding is compiled the ciphertext C for needing to encode by the corresponding encoded mode that data analysis phase determines
Code becomes specific bit string, and group becomes mode indicators+character count indicator+encoded digital information bit stream form, mentions
Supplied a variety of Data Coding Scheme for being selected according to practical application, including ECI mode, figure pattern, octet mode,
Alphanumeric mode and Chinese mode, each mode have its corresponding character set and corresponding coding to be supported to advise
Then table;
Step 3.3, Error Correction of Coding and final code word construction, generate error correction code word for coded data, to provide two-dimentional code symbol
Number certain error resilience performance itself, two dimensional code standard has formulated the error correcting capability of four grades altogether, in the process of Error Correction of Coding
In, need to be arranged reasonable error-correction level in conjunction with the symbol version of selection and application.According to the version and level of error correction of selection,
Code word data is subjected to piecemeal according to relevant criterion, it is corresponding to generate the piecemeal with error correction algorithm respectively to each small piecemeal
Error correction code word, and be appended to after corresponding code word data, form the final code word sequence of symbol;
Step 3.4, module is arranged in a matrix, and the work which mainly does is blank corresponding with symbol version
In square matrices, complete expression and its location layout of functional graphic and sign character, the position of functional graphic and structure oneself
Fixed via standard, dark module be set on the corresponding position in functional graphic area according to standard, for sign character, with length,
Height is respectively the regular rectangular shape module array of 2 and 4 block sizes to indicate, but if in the symbols such as itself and positioning pattern
Functional graphic when suffering close, then be converted into irregular form to reduce mutual interference, layout in a matrix
It since the bottom right of data-encoding area, turns left from the right side, is from top to bottom filled with 2*4 rule module;
Step 3.5, exposure mask operates the non-functional graph area of symbol, the reference pattern provided with standard and its life
XOR operation is carried out to the matrix of previous step at condition, all results are evaluated, it is best to select effect, so that symbol
In uneven color module be evenly distributed as far as possible, generate two dimensional code;
Wherein, payment process is,
Step 1, code reader, which scans the two-dimensional code, obtains ciphertext C, and decrypts, and uses private key (k2, X) and original to decrypt acquisition
Data M, then,
Step 2, initial data M is converted into telephone number, identification card number, payment cipher, payment information, is paid
Operation.
It is all that ensure that transaction using encrypted data, therefore greatly during two dimensional code is generated to payment
Safety.
Payment way be broadly divided into uniformly supplement with money with two kinds of self-recharging, be applicable in different scenes respectively, uniformly supplement master with money
Being used for unit batch is that visitor supplements with money, and self-recharging only supports visitor individual to supplement with money;
Code reader scanning two dimension payment code is consumed, and administrator's consumption statistic is divided into two kinds and checks that mode, one kind are to visit
The consumption details of objective querying individual, it is another to check the amount of money that every code reader is taken in.
Embodiment described above only expresses one embodiment of the present invention, and but it cannot be understood as right
The limitation of the scope of the invention.It should be pointed out that for those of ordinary skill in the art, not departing from present inventive concept
Under the premise of, various modifications and improvements can be made, and these are all within the scope of protection of the present invention.
Claims (10)
1. a kind of artificial intelligence visitor management system with face recognition, including intelligent terminal, external network server, Intranet service
Device, access control equipment, vehicle management equipment and payment devices, visitor access external network server by intelligent terminal, are reserved
Application, examining department examines the reservation of visitor by network server, and approval results are fed back to outer net clothes respectively
Business device and intranet server, external network server generate reservation feedback according to approval results and are sent to intelligent terminal, intranet server
It is connect respectively with access control equipment, vehicle management equipment and payment devices, if reservation is fed back through, access control equipment will allow to hold
The visitor for having the reservation to feed back enters Office Area, and the vehicle that vehicle management equipment allows to hold reservation feedback enters parking lot,
The visitor for allowing to hold reservation feedback is carried out payment activity by payment devices;Access control equipment includes that monitoring camera, fingerprint are adopted
Acquisition means, bio-identification access controller and electronic lock, the real-time visitor's face-image of monitoring camera dynamic acquisition, fingerprint are adopted
Collect equipment and acquire real-time visitor's fingerprint, bio-identification access controller can be to real-time visitor's face-image and real-time visitor's fingerprint
It is identified, and carries out matching comparison, if after the two all compares successfully, opening electronic lock, allow visitor by gate inhibition, if
The two compares repeatedly failure, can starting alarm in real time.
2. a kind of artificial intelligence visitor management system with face recognition according to claim 1, it is characterised in that: face
The matching comparison process of portion's image is as follows,
Step 1, size normalizes, specifically:
Step 1.1,27 facial fiducial points are obtained, be respectively as follows: right pupil, left pupil, nose, the right corners of the mouth, the left corners of the mouth, outside right eyebrow
Endpoint in endpoint, right eyebrow, right eye socket of the eye the outer end point, right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point, endpoint, left eyebrow inner end in right eye socket of the eye
Point, left eyebrow the outer end point, endpoint, left eye socket of the eye upper extreme point, left eye socket of the eye lower extreme point, left eye socket of the eye the outer end point, nasion portion right end in left eye socket of the eye
Under point, nasion portion left end point, wing of nose right part, wing of nose left part, wing of nose right endpoint, wing of nose left end point, upper lip upper extreme point, upper lip
Endpoint, lower lip upper extreme point, lower lip lower extreme point;
Said reference point is by the extracting section of face most feature, to improve the precision of identification;
Step 1.2, it is determined in right eye according to endpoint in right eye socket of the eye the outer end point, right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point, right eye socket of the eye
Heart point determines left eye central point according to endpoint, left eye socket of the eye upper extreme point, left eye socket of the eye lower extreme point, left eye socket of the eye the outer end point in left eye socket of the eye, seeks
The center of face of right eye central point and left eye central point point is looked for, using this center of face point as image origin, rotates face-image,
The line for being linked to be right eye central point and left eye central point is in horizontal position, to rectify face-image;
Step 1.3, facial ratio characteristic is combined according to facial fiducial point, facial major part is cut into, it thereafter will cutting
Face-image afterwards is the image of uniform sizes size by scale transformation, hair, background redundancy except elimination facial parts
Information;
Step 2, gray scale normalization,
Step 3, histogram is constructed,
Step 4, similarity is calculated,
Real-time visitor's face-image and reservation mug shot is uploaded by step 1-3 to be handled to obtain respective histogram point
Amount, respectivelyWithBy comparing two histogram components, to calculate similarity between the two, specific calculate is constituted such as
Under:
Successful match is thought if D is less than 0.3, otherwise it fails to match.
3. a kind of artificial intelligence visitor management system with face recognition according to claim 2, it is characterised in that step
Rapid 1.4 specifically:
Step 1.4.1 calculates the distance d of right eye socket of the eye upper extreme point, right eye socket of the eye lower extreme point;
Step 1.4.2 calculates center of face point and the left and right endpoint midpoint distance l of the wing of nose;
Step 1.4.3 calculates eye socket distance ratio r=d/l, ratio characteristic of the eye socket distance than having reacted face;
Step 1.4.4, according to eye socket distance than as origin uniform zoom and cutting face-image using the left and right endpoint midpoint of the wing of nose.
4. a kind of artificial intelligence visitor management system with face recognition according to claim 2, it is characterised in that step
Rapid 2 specifically:
Step 2.1, gray scale adjusts,
If size is the grayscale image of W × H, gray matrix function is I (i, j), and I (i, j) represents the intensity of each pixel, is adjusted
Gray matrix function after whole be S (i, j) then,
S (i, j)=c × cos θ × log (1+I (i, j)),
Wherein, c is constant, and θ is shooting angle,
Step 2.2, matrix function is constructed,
Normalization matrix function is G (i, j), then,
Wherein μ is that gray average is before normalizing, and normalization front difference is σ, then, gray average is μ after normalization0, normalization side
Difference is σ0;
Step 3 specifically:
Step 3.1, the space λ is constructed, the gradient magnitude T (i, j) and deflection A (i, j) of each pixel, enhancing profile letter are calculated
Breath,
Wherein, Tx(i, j) and Ty(i, j) respectively represents the gradient magnitude on both horizontally and vertically,
Step 3.2, centered on each facial fiducial point, the pixel square region that 27 sizes are 100 × 100 is constructed, building
The histogram component function h of every rowj(i, j)=A (i, j) × T (i, j), i=1,2......100, j=1,2......100, directly
Square component is hj=[hj(1,j),hj(2,j)......hj(i,j),......hj(100, j)], construct the straight of each square region
Square component Hk=[h1,h2...hj...h100], k=1,2......27.
5. a kind of artificial intelligence visitor management system with face recognition according to claim 1, it is characterised in that: pre-
The detailed process about applied are as follows: visitor logs in the reservation system in external network server by intelligent terminal and looked into according to working demand
The reservation situation of visiting department required for asking, after choosing visiting department, visitor checks the recent reservation situation of the visiting department, with
Just the suitable time is selected to be reserved, visitor fill in reservation information need to be by name, telephone number, identification card number, visiting portion
Door, visiting time fill in strictly according to the facts and mug shot and fingerprint are uploaded to reservation system, fill in same administrative staff or vehicle according to demand
Information.
6. a kind of artificial intelligence visitor management system with face recognition according to claim 1, it is characterised in that: raw
At the detailed process of the reservation feedback are as follows: for visitor after reservation application, examining department receives access by reservation system immediately
Application, waits examining department's examination & approval, and the approval results of examining department are generated reservation feedback, pass through wechat or short message by reservation system
Platform is sent to the intelligent terminal of visitor;Reservation feedback includes the name of visitor, telephone number, identification card number, visiting department, arrives
Visit time, mug shot, fingerprint, information of vehicles, subscription state, permission.
7. a kind of artificial intelligence visitor management system with face recognition according to claim 6, it is characterised in that: pre-
About apply for that the pending time is 30 minutes, if not processed in 30 minutes, reservation system will be carried out certainly application according to reservation situation
The dynamic intelligent terminal given an written reply and visitor is sent to by wechat or SMS platform;Reservation feedback will be by automatic input intranet server
Database in, related personnel and visitor can pass through platform or wechat query-reservation application at any time.
8. a kind of artificial intelligence visitor management system with face recognition according to claim 6, it is characterised in that: pre-
About system includes backstage authority management module, and examining department combines reservation feedback to carry out permission to visitor by authority management module
Setting, allows external staff to exempt to reserve within a certain period of time, operates in the period in access control equipment for exempting from reservation
The open multiple right of access of staff, and the visiting department for exempting from the external staff of reservation and visiting number are done accordingly
Record and count typing intranet server database in.
9. a kind of artificial intelligence visitor management system with face recognition according to claim 1, it is characterised in that: vehicle
Management equipment includes Car license recognition camera, vehicle room entry/exit management device and channel milling train, and vehicle room entry/exit management device knows license plate
The information of vehicles of other cameras capture is compared with the license board information in the database of intranet server, after comparing successfully, by vehicle
Trade mark code, type of vehicle, vehicle color, entry time, mode of operation, alarm condition and license plate figure are uploaded to intranet server
Database in, vehicle room entry/exit management device control channel milling train is let pass, if compared repeatedly, failure, vehicle room entry/exit management device if can
Automatic upload information is to administrative staff;Car license recognition cameras capture information of vehicles and with the license plate in the database of intranet server
The license plate figure including Car license recognition cameras capture is compared after image recognition and in the database of intranet server in information
License plate number compare, type of vehicle compares and vehicle color compares, administrative staff pass through the data of vehicle room entry/exit management device
Visualization interface such as is inquired the information of vehicles in the database of intranet server, is increased newly, being modified, being deleted at the operation.
10. a kind of artificial intelligence visitor management system with face recognition according to claim 1, it is characterised in that:
Payment management includes code reader and top-up machines, and top-up machines generate unique according to typing visitor name, telephone number, identification card number
Visitor's two dimension payment code, is used for identification, visitor is supplemented with money and paid by top-up machines using the payment code, supplemented with money
Mode be broadly divided into uniformly supplement with money with two kinds of self-recharging, be applicable in different scenes respectively, uniformly supplement with money and be mainly used for unit batch
Amount is that visitor supplements with money, and self-recharging only supports visitor individual to supplement with money, and code reader scanning two dimension payment code is consumed, and is managed
Member's consumption statistic is divided into two kinds and checks mode, a kind of consumption details for visitor's querying individual, and another kind is to check every barcode scanning
The amount of money that device is taken in.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810334327.0A CN109300200A (en) | 2018-04-15 | 2018-04-15 | A kind of artificial intelligence visitor management system with face recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810334327.0A CN109300200A (en) | 2018-04-15 | 2018-04-15 | A kind of artificial intelligence visitor management system with face recognition |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109300200A true CN109300200A (en) | 2019-02-01 |
Family
ID=65167527
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810334327.0A Pending CN109300200A (en) | 2018-04-15 | 2018-04-15 | A kind of artificial intelligence visitor management system with face recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109300200A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111126198A (en) * | 2019-12-11 | 2020-05-08 | 中山大学 | Pedestrian re-identification method based on deep representation learning and dynamic matching |
CN112714275A (en) * | 2019-10-24 | 2021-04-27 | 上海宗保科技有限公司 | Automatic monitoring security system and use method thereof |
CN113066214A (en) * | 2021-03-26 | 2021-07-02 | 深圳市博盛科电子有限公司 | Access control system based on 5G network remote monitoring |
CN114170716A (en) * | 2021-10-29 | 2022-03-11 | 上海晨晓电子科技有限公司 | Nuclear power access control system and method thereof |
CN114924823A (en) * | 2022-05-27 | 2022-08-19 | 中国银行股份有限公司 | Language management method of self-service teller machine, related device and storage medium |
WO2022261974A1 (en) * | 2021-06-18 | 2022-12-22 | 京东方科技集团股份有限公司 | Information management method, apparatus, system, and storage medium |
CN116358902A (en) * | 2023-06-02 | 2023-06-30 | 中国第一汽车股份有限公司 | Vehicle function testing method and device, electronic equipment and storage medium |
-
2018
- 2018-04-15 CN CN201810334327.0A patent/CN109300200A/en active Pending
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112714275A (en) * | 2019-10-24 | 2021-04-27 | 上海宗保科技有限公司 | Automatic monitoring security system and use method thereof |
CN111126198A (en) * | 2019-12-11 | 2020-05-08 | 中山大学 | Pedestrian re-identification method based on deep representation learning and dynamic matching |
CN111126198B (en) * | 2019-12-11 | 2023-05-09 | 中山大学 | Pedestrian re-identification method based on deep representation learning and dynamic matching |
CN113066214A (en) * | 2021-03-26 | 2021-07-02 | 深圳市博盛科电子有限公司 | Access control system based on 5G network remote monitoring |
WO2022261974A1 (en) * | 2021-06-18 | 2022-12-22 | 京东方科技集团股份有限公司 | Information management method, apparatus, system, and storage medium |
CN114170716A (en) * | 2021-10-29 | 2022-03-11 | 上海晨晓电子科技有限公司 | Nuclear power access control system and method thereof |
CN114924823A (en) * | 2022-05-27 | 2022-08-19 | 中国银行股份有限公司 | Language management method of self-service teller machine, related device and storage medium |
CN116358902A (en) * | 2023-06-02 | 2023-06-30 | 中国第一汽车股份有限公司 | Vehicle function testing method and device, electronic equipment and storage medium |
CN116358902B (en) * | 2023-06-02 | 2023-08-22 | 中国第一汽车股份有限公司 | Vehicle function testing method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108335396A (en) | A kind of artificial intelligence visitor management system with Car license recognition | |
CN109300200A (en) | A kind of artificial intelligence visitor management system with face recognition | |
CN108629873A (en) | A kind of artificial intelligence visitor management system with Quick Response Code payment | |
US7158657B2 (en) | Face image recording system | |
JP4246154B2 (en) | Biometric authentication method | |
JP2007272320A (en) | Entry management system | |
CN106981016A (en) | A kind of remote self-help real name buys the method and system of phonecard | |
CN103793990A (en) | Invoice verification method and system | |
CN1860724A (en) | Method for identification | |
CN106710032B (en) | Entrance guard opening method and device | |
CN108986342B (en) | Face recognition locker system based on cloud computing platform | |
WO2012124115A1 (en) | Biological information acquisition device, biological information comparison device, and program | |
CN114863506A (en) | Method, device and system for verifying access permission and identity authentication terminal | |
CN106296949B (en) | A kind of access control system and method based on H264 video decoding techniques | |
CN113421079A (en) | Borrowing and returning shared charger baby method based on shared charger baby rental cabinet | |
CN112258707A (en) | Intelligent access control system based on face recognition | |
CN111932754B (en) | Laboratory entrance guard verification system and verification method | |
CN106710033B (en) | Entrance guard opening method and device | |
CN115620117A (en) | Face information encryption method and system for network access authority authentication | |
Fazilov et al. | Access control system based on facial image analysis | |
CN115982687A (en) | User identity verification system for data operation and maintenance management platform | |
Lekshmi et al. | Source camera identification of image for forensic analysis using sensor fingerprints | |
CN117320008B (en) | 5G message tax system and processing method thereof | |
CN115527293B (en) | Method for opening door by security chip based on human body characteristics and security chip device | |
CN116052313B (en) | Intelligent secret cabinet control method, device, equipment and storage medium |
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 | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20200102 Address after: Room 201, unit 3, building 16, No. 3, Baisha Road, Tangjiawan, Xiangzhou District, Zhuhai City, Guangdong Province Applicant after: Chen Huiping Address before: 519085 Unit RB09, First Floor, No. 1 Teaching Area, Software Park Road, Tangjiawan Town, Zhuhai High-tech Zone, Guangdong Province Applicant before: ZHUHAI HUAQING INNOVATION TECHNOLOGY Co.,Ltd. |
|
TA01 | Transfer of patent application right | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190201 |
|
WD01 | Invention patent application deemed withdrawn after publication |