CN111932754A - Laboratory entrance guard verification system and verification method - Google Patents

Laboratory entrance guard verification system and verification method Download PDF

Info

Publication number
CN111932754A
CN111932754A CN202010558650.3A CN202010558650A CN111932754A CN 111932754 A CN111932754 A CN 111932754A CN 202010558650 A CN202010558650 A CN 202010558650A CN 111932754 A CN111932754 A CN 111932754A
Authority
CN
China
Prior art keywords
face
data
identification
entrance guard
identification degree
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010558650.3A
Other languages
Chinese (zh)
Other versions
CN111932754B (en
Inventor
迟海鹏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Dynaflow Experiment Technology Co Ltd
Original Assignee
Beijing Dynaflow Experiment Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Dynaflow Experiment Technology Co Ltd filed Critical Beijing Dynaflow Experiment Technology Co Ltd
Priority to CN202010558650.3A priority Critical patent/CN111932754B/en
Publication of CN111932754A publication Critical patent/CN111932754A/en
Application granted granted Critical
Publication of CN111932754B publication Critical patent/CN111932754B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G07C9/30—Individual registration on entry or exit not involving the use of a pass
    • G07C9/32—Individual registration on entry or exit not involving the use of a pass in combination with an identity check
    • G07C9/37—Individual registration on entry or exit not involving the use of a pass in combination with an identity check using biometric data, e.g. fingerprints, iris scans or voice recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/56—Extraction of image or video features relating to colour
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • 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
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03F—AMPLIFIERS
    • H03F3/00—Amplifiers with only discharge tubes or only semiconductor devices as amplifying elements
    • H03F3/68—Combinations of amplifiers, e.g. multi-channel amplifiers for stereophonics

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Power Engineering (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)

Abstract

The invention belongs to the technical field of laboratory equipment, and particularly relates to a laboratory entrance guard verification system and a laboratory entrance guard verification method with higher identification precision. The method comprises the following steps: the face contour recognition module is used for recognizing the face contour data and confirming the identification degree LB of the face contour data; the face facial features recognition module is used for recognizing the position and shape data of facial features and confirming the identification degree RB of the facial features data; the face complexion identification module is used for identifying the complexion information of the face and confirming the face complexion data identification degree FB; and the identification degree comprehensive processing module is used for obtaining the laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB. The algorithm is transplanted to an embedded platform based on ARM and FPGA architectures of a ZYNQ platform, the algorithm is conveniently improved and combined with an FPGA unit, and the identification accuracy and the real-time performance are met.

Description

Laboratory entrance guard verification system and verification method
Technical Field
The invention belongs to the technical field of laboratory equipment, and particularly relates to a laboratory entrance guard verification system and a laboratory entrance guard verification method with higher identification precision.
Background
With the development of the scientific and technological society and the increase of the world economy, the development of an intelligent laboratory is closely related to the experimental teaching of university, and the intelligent degree of the laboratory also determines the experimental quality of students. However, even in the intelligent laboratory, the problem that the experimental article is stolen by the outside personnel still exists, and the theft rate cannot be improved in the relatively old laboratory because the door is directly opened for free access. Aiming at the problems, the identity recognition technology becomes a research hotspot at home and abroad, and each country is dedicated to developing an intelligent access control technology to improve the safety of a laboratory, and the development of the technology ensures the safety and benefits of the country and the school.
The entrance guard system as one modern safety management system combines automatic machine identifying technology and modern safety management measures into one integral, and relates to electronic technology, mechanical automation, computer technology, communication technology, biological identification technology and other new technologies. The access control system is an effective measure for solving the safety problems of various experimental places and external entrances and exits. In a laboratory, many scenes such as an intelligent laboratory cabin, a laboratory and the like can not be opened without an access control system.
Early access control system is through the form of doorbell, contacts laboratory administrator through the entrance guard, and laboratory administrator controls opening of entrance guard, and this kind of access control system's advantage lies in with low costs, easily control. However, the access control system only aims at external personnel, a laboratory administrator still needs to use a key to open the door, the use is very inconvenient, and the early access control system does not have any automatic identification technology and only improves the safety through the control direction. With the development of the sensing technology and the rapid development of the electronic technology market, the entrance guard system enters a rapid development period, for example, the induction card type entrance guard system is commercialized, so that the convenience and the like of the entrance guard system are improved, and because the product cost is low, many similar products have a certain market in the face of a lower-end scene. However, the access control card is still an article to be carried, which has certain limitations, and once the access control card is lost or damaged, great troubles are brought. Therefore, in recent decades, with the gradual maturity of biometric identification technology and the rapid development of artificial intelligence, the entrance guard system has also been developed from the early simple control to a system capable of performing automatic identification by using human body biometrics. At present, the entrance guard system has been developed to the maturity stage, and the safety problems of the entrances and exits of laboratories in various scenes are favorably improved due to the vigorous development of the fingerprint entrance guard system, the iris entrance guard system, the voiceprint recognition entrance guard system, the face recognition entrance guard system and the like. Fingerprint access control system and palm print discernment access control system are comparatively convenient for the access control card, but because the information quantity is less and easily forged, therefore factor of safety is not high enough. Iris access control system need not consider the sanitary problem that frequent contact brought for fingerprint access control system, and because be human iris information as the password, be difficult to forge, but because the particularity of its work, need the people to press close to equipment during verification, because people's height difference is great, lead to equipment interactivity not very good, dedicated image acquisition equipment cost is high and the installation is comparatively troublesome, and the cost is higher can't promote on a large scale. The voiceprint recognition access control system performs identity recognition on a user by utilizing voiceprint information stored in advance, the system cost is relatively low, but due to the limitation of the voiceprint information, the accuracy is not high enough and the voiceprint recognition access control system is easy to forge.
Face identification access control system has non-contact's advantage equally, and better, the convenience is stronger for iris identification system's human-computer interaction, because face image's uniqueness is difficult to be counterfeited equally, factor of safety is high, and is high for use in the iris system and discernment accuracy, can be fast and accurate completion identity's identification verification. Not only an access control system, but also a face recognition technology as a biological feature recognition technology has huge potential application scenes. Face recognition technology has been a popular topic in the fields of computer vision, machine learning and the like in recent years.
Disclosure of Invention
The invention aims to provide a laboratory entrance guard verification system with higher identification precision.
The invention also aims to provide a laboratory entrance guard verification method.
The purpose of the invention is realized as follows:
a laboratory access verification system comprising:
the face contour recognition module is used for recognizing the face contour data and confirming the identification degree LB of the face contour data;
the face facial features recognition module is used for recognizing the position and shape data of facial features and confirming the identification degree RB of the facial features data;
the face complexion identification module is used for identifying the complexion information of the face and confirming the face complexion data identification degree FB;
the identification degree comprehensive processing module is used for obtaining a laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB;
if the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
The face contour recognition module recognizes face contour data and confirms the face contour data identification degree LB, and comprises the following steps:
(1.1) acquiring n characteristic points of all human face contours capable of being recognized, which are stored in the laboratory entrance guard verification system, and calculating the fluctuation level of each characteristic point of the human face contour;
Figure BDA0002545285010000021
wherein the content of the first and second substances,
Figure BDA0002545285010000022
as an intermediate parameter, Hbdi,jFor the jth feature point, σ, of the face contour of the ith person stored in the lab-gate verification systemjThe fluctuation level of the jth characteristic point of the face contour is m, the total amount of faces stored in the laboratory entrance guard verification system is i-1, 2, 3.. m, and j-1, 2,3 … n;
(1.2) calculating an adjusting coefficient of each feature point of the human face contour;
Figure BDA0002545285010000023
Hjadjusting coefficients of j characteristic points of the human face contour;
(1.3) acquiring n characteristic points of the face contour of the person to be recognized;
(1.4) calculating the difference between each characteristic point of the face contour of the person to be recognized and all the face contours of the recognizable persons stored in the laboratory entrance guard verification system;
Figure BDA0002545285010000024
Figure BDA0002545285010000025
the difference between the jth characteristic point of the face contour of the person to be recognized and the face contour of the jth characteristic point of the ith person stored in the laboratory entrance guard verification system is obtained;
(1.5) calculating the face contour data identification degree LB:
Figure BDA0002545285010000031
Figure BDA0002545285010000032
Figure BDA0002545285010000033
LB＝max(LBi,LB2,LB3,…,LBi…LBj)
wherein the content of the first and second substances,
Figure BDA0002545285010000034
LBiis an intermediate parameter;
and evaluating the conformity of the face contour curve of the identified person according to the size of the face contour data identification degree LB, wherein the smaller LB is, the lower the conformity of the identified person is.
The facial features recognition module recognizes the position and shape data of facial features and confirms the recognition degree RB of the facial features data, and the facial features recognition module comprises the following steps:
firstly, acquiring the positions of five sense organs of a human face:
(2.1) carrying out convolution processing on the face image by using a 2D Gaussian filtering template and an original image to eliminate the noise of the face image;
(2.2) carrying out gray processing on the face image, and calculating the growth rate of pixel values along the horizontal left direction and the horizontal lower direction at each position of the gray image;
(2.3) calculating a comprehensive growth rate of each position;
Figure BDA0002545285010000035
wherein f (x, y) is the comprehensive growth rate of the position with the face image coordinate of (x, y), and HzIs the growth rate of the pixel value of the position with coordinates (x, y) of the face image along the horizontal left direction, HxThe growth rate of the pixel value of the position with the coordinate (x, y) of the face image along the horizontal downward direction;
(2.4) calculating a growth direction for each location;
Figure BDA0002545285010000036
wherein, the coordinate of the theta (x, y) face image is an included angle between the position of (x, y) along the horizontal left direction and the growth direction;
(2.5) traversing the whole face image, comparing whether the value of the pixel point at each position of the face image after gray scale is larger than the values of the pixel points at the front position and the rear position in the increasing direction of the position, if so, not changing the value of the pixel point at the position, otherwise, assigning the value of the pixel point at the position as 0;
(2.6) taking the point of the image processed in the step (2.5) with the value of the pixel point larger than a preset value T1 as a position point of the five sense organs, so as to obtain the coordinate positions of the five sense organs;
and then, taking the reciprocal of the Euclidean distance between the obtained coordinate position of the five sense organs and the preset coordinate position of the standard five sense organs as the data identification degree RB of the five sense organs of the face.
The face complexion identification module is used for identifying complexion information of a face and confirming the face complexion data identification degree FB; the method comprises the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{Z，fs_min,fs_max}
wherein K is a skin color data set, Z is the brightness of the extracted face skin color, fs _ min is the minimum value of the pixels of the extracted face skin color, and fs _ max is the maximum value of the pixels of the extracted face skin color;
(3.2) obtaining the values of pixel points of the person to be identified, and calculating the weight of each pixel point value;
Figure BDA0002545285010000041
W_xsx,ythe weight value of the gray value of the pixel point of the position with the face image coordinate being (x, y),
H_xsx,ythe gray value of a pixel point of a position with the face image coordinate of (x, y);
h _ xs is the gray value of pixel points at all positions of the face image;
(3.3) calculating the belonged value of the skin color;
Figure BDA0002545285010000042
wherein G _ fsx,yThe probability of the skin color of the position with the coordinates of (x, y) of the face image is shown as cov, the covariance matrix after covariance of H _ xs by columns is shown as H _ xs,
Figure BDA0002545285010000043
to average the H xs column by column,
Figure BDA0002545285010000044
to pair
Figure BDA0002545285010000045
Transposition is carried out;
(3.4) confirming the identification degree FB of the face complexion data;
Figure BDA0002545285010000046
wherein, the (x, y) ∈ ν is the value of the coordinates (x, y) of the face image, and the value is all coordinate points of the face image.
Obtaining the laboratory entrance guard verification identification TB according to the following formula:
Figure BDA0002545285010000051
wherein, a and b are preset coefficients, the preset values are 1 +/-0.02, max (LB) is the maximum value that LB can take, max (RB) is the maximum value that RB can take, and max (FB) is the maximum value that FB can take.
If the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
The face contour recognition module recognizes face contour data and confirms the face contour data identification degree LB, and comprises:
(1.1) inquiring the face outline data index X of the identified personmWherein m is 1,2,3, wherein XmFirst to third face contour models of a recognized person;
(1.2) calculating the face contour index adjustment coefficient R of the identified personi；
R1＝(X1-p1)/p1；R2＝(X2-p2)/p2；R3＝(X3-p3)/p3；
p·1Is X1The face contour ratio adjustment coefficient of (1); p is a radical of2Is X2The face contour ratio adjustment coefficient of (1); p is a radical of3Is X3The face contour ratio adjustment coefficient of (1);
(1.3) calculating a weighted face contour ratio adjustment coefficient P of the recognized person:
P＝1+R1X1+R2X2+R3X3；
(1.4) calling the average value V of the face contour model of the face contour of the identified person to obtain a value multiplier U of the face contour of the identified person,
U＝VP；
(1.5) calling all the collected face contour line data v of the identified personiI ∈ (1, 2, …, n) confirms the face contour data identification degree LB:
Figure BDA0002545285010000052
and evaluating the conformity of the face contour curve of the identified person through the size of the LB, wherein the smaller the LB, the lower the conformity of the identified person is.
The identification RB for confirming the facial feature data comprises the following steps:
(2.1) let the coordinates of the facial feature pixel iv be (X)iv，Yiv) Let the coordinates of g pixels of facial features not matching the coordinates be (X)1，Y1)，(X2，Y2)…(Xg，Yg)，g≥3；
(2.2) the Euclidean distance between the facial five sense organ pixels which do not coincide with the coordinates and the associated m facial five sense organ pixels of the person is as follows:
Figure BDA0002545285010000053
the above equation is derived:
Figure BDA0002545285010000054
the formula is calculated through a simplified structure to obtain the adjustment coordinates of facial feature pixels which are not matched with the coordinates:
Figure BDA0002545285010000055
Figure BDA0002545285010000061
Figure BDA0002545285010000062
Figure BDA0002545285010000063
(2.3) setting the deviation distance r of the facial feature pixel iu and a standard storage facial feature pixel model, wherein the communication error of a data acquisition image system is r0(ii) a The average power value of the face acquisition system is P (r)0) (ii) a The image receiving power of the facial five-organ pixels which are not matched with the coordinates is P (r); then:
Figure BDA0002545285010000064
alpha is an image acquisition path loss index, and the deviation distance r between the facial feature pixel iu and a standard stored facial feature pixel model is obtained through calculation;
(2.4) calculating and confirming the identification degree RB of the facial feature data;
Figure BDA0002545285010000065
d0is the detection distance, X, of the system from the acquired faceσIs a positive distribution of random values with a standard deviation of [3, 6 ]]A (i) data acquisition loss rate for facial features; h0The original signal strength of the original facial features data is collected for the system.
The confirmation of the identification degree FB of the face skin color data comprises the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{k0,k1,k2,…,kv}；
kvstoring data for a skin tone database; k is a radical of0As standard skin tone data, k0The deviation data of (a) includes:
me(k0)＝{k01,k02,k02,…,k0t}；
t is the number of deviation data;
(3.2) estimating K and measured skin tone data sqThe color difference between the two images is obtained,
Figure BDA0002545285010000066
Rele(W,G(me(k0) ) represents the correlation degree of the skin color data corresponding to the face skin color database and the measured skin color data, G(s)q) A set of correlation values representing measured skin color data obtained by the acquisition;
(3.3) confirming the identification degree FB of the face complexion data
Figure BDA0002545285010000067
Wei denotes K and CSCorrelation between, D (K, C)S) Represents K and CSThe color difference value between them.
The obtained laboratory entrance guard verification identification TB is as follows:
Figure BDA0002545285010000071
body identification coefficient P ═ alpha (LB + RB)t
Weak identification coefficient AOC is 24.2 XFB2。
A laboratory entrance guard verification method comprises the following steps:
recognizing the face contour data and confirming the identification degree LB of the face contour data, comprising the following steps:
(1.1) acquiring n characteristic points of all human face contours capable of being recognized, which are stored in a laboratory entrance guard verification system, and calculating the fluctuation level of each characteristic point of the human face contour;
Figure BDA0002545285010000072
wherein the content of the first and second substances,
Figure BDA0002545285010000073
as an intermediate parameter, Hbdi,jFor the jth feature point, σ, of the face contour of the ith person stored in the lab-gate verification systemjThe fluctuation level of the jth characteristic point of the face contour is m, the total amount of faces stored in the human laboratory entrance guard verification system is i-1, 2, 3.. m, and j-1, 2,3 … n;
(1.2) calculating an adjusting coefficient of each feature point of the human face contour;
Figure BDA0002545285010000074
Hjadjusting coefficients of j characteristic points of the human face contour;
(1.3) acquiring n characteristic points of the face contour of the person to be recognized;
(1.4) calculating the difference between each characteristic point of the face contour of the person to be recognized and all the face contours of the recognizable persons stored in the laboratory entrance guard verification system;
Figure BDA0002545285010000075
Figure BDA0002545285010000076
the difference between the jth characteristic point of the face contour of the person to be recognized and the face contour of the jth characteristic point of the ith person stored in the laboratory entrance guard verification system is obtained;
(1.5) face contour data identification degree LB:
Figure BDA0002545285010000077
Figure BDA0002545285010000078
Figure BDA0002545285010000079
LB＝max(LBi,LB2,LB3,…,LBi…LBj)
wherein the content of the first and second substances,
Figure BDA0002545285010000081
LBiis an intermediate parameter;
and evaluating the conformity of the face contour curve of the identified person through the size of the LB, wherein the smaller the LB, the lower the conformity of the identified person is.
Confirming the identification degree RB of the facial feature data, comprising the following steps:
firstly, the positions of five sense organs of the human face are obtained
(2.1) carrying out convolution processing on the face image by using a 2D Gaussian filtering template and an original image to eliminate the noise of the face image;
(2.2) carrying out gray processing on the face image, and increasing the pixel value growth rate of each position of the grayed image along the horizontal left direction and the horizontal lower direction;
(2.3) calculating a comprehensive growth rate of each position;
Figure BDA0002545285010000082
wherein f (x, y) is the comprehensive growth rate of the position with the face image coordinate of (x, y), and HzIs the growth rate of the pixel value of the position with coordinates (x, y) of the face image along the horizontal left direction, HxThe growth rate of the pixel value of the position with the coordinate (x, y) of the face image along the horizontal downward direction;
(2.4) calculating a growth direction for each location;
Figure BDA0002545285010000083
wherein, the coordinate of the theta (x, y) face image is an included angle between the position of (x, y) along the horizontal left direction and the growth direction;
(2.5) traversing the whole face image, comparing whether the value of a pixel point after the gray level of the face image is larger than the values of pixel points at the front position and the rear position in the growth direction of the face image, and setting the value of the pixel point not to be 0;
(2.6) taking the point of the image processed in the step (2.5) with the value of the pixel point larger than a preset value T1 as a position point of the five sense organs, so as to obtain the coordinate positions of the five sense organs;
and then, taking the reciprocal of the Euclidean distance between the obtained coordinate position of the five sense organs and the preset coordinate position of the standard five sense organs as the data identification RB of the five sense organs of the human face.
Confirming the identification degree FB of the face complexion data, comprising the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{Z，fs_min,fs_max}
wherein K is a skin color data set, Z is the brightness of the extracted face skin color, fs _ min is the minimum value of the pixels of the extracted face skin color, and fs _ max is the maximum value of the pixels of the extracted face skin color;
(3.2) obtaining the values of pixel points of the person to be identified, and calculating the weight of each pixel point value;
Figure BDA0002545285010000091
W_xsx,ythe weight value of the gray value of the pixel point of the position with the face image coordinate being (x, y),
H_xsx,ythe gray value of a pixel point of a position with the face image coordinate of (x, y);
h _ xs is the gray value of pixel points at all positions of the face image;
(3.3) calculating the belonged value of the skin color;
Figure BDA0002545285010000092
wherein G _ fsx,yThe probability of the skin color of the position with the coordinates of (x, y) of the face image is shown as cov, the covariance matrix after covariance of H _ xs by columns is shown as H _ xs,
Figure BDA0002545285010000093
to average the H xs column by column,
Figure BDA0002545285010000094
to pair
Figure BDA0002545285010000095
Transposition is carried out;
(3.4) confirming the identification degree FB of the face complexion data
Figure BDA0002545285010000096
Wherein, the (x, y) ∈ ν is the value of the coordinates (x, y) of the face image, and the value is all coordinate points of the face image.
The identification degree comprehensive processing module is used for obtaining a laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB;
the obtained laboratory entrance guard verification identification TB is as follows:
Figure BDA0002545285010000097
wherein, a and b are preset coefficients, the preset values are 1 +/-0.02, max (LB) is the maximum value that LB can take, max (RB) is the maximum value that RB can take, and max (FB) is the maximum value that FB can take.
If the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
A laboratory entrance guard verification system, a human face contour recognition module comprises a recognition circuit,
the identification circuit includes: a first resistor R1, a second resistor R2, a third resistor R3, a fourth resistor R4, a fifth resistor R5, a sixth resistor R6, a seventh resistor R7, an eighth resistor R8, a ninth resistor R9, a tenth resistor R10, a first capacitor C1, a second capacitor C2, a third capacitor C3, a fourth capacitor C4, a fifth capacitor C5, a first NPN transistor Q1, a second NPN transistor Q2, a third NPN transistor Q3, a fourth NPN transistor Q4, a diode L, a first operational amplifier B1, a second operational amplifier B2, a third operational amplifier B3, a fourth operational amplifier B4, a first amplifier a1, a second amplifier a2, a discriminator, a power source VCC, and a ground GND;
the power supply VCC is respectively connected with the input ends of the identifier, the second resistor R2, the fifth resistor R5 and the first capacitor C1, the output end of the second resistor R2 is respectively connected with the input end of the third resistor R3 and the collector of the first NPN transistor Q1, the output end of the first capacitor C1 is connected with the base of the first NPN transistor Q1, and the emitter of the first NPN transistor Q1 is connected with the input end of the first resistor R1;
the output end of the third resistor R3 is connected with the positive input end of a first operational amplifier B1, the inverting input end of the first operational amplifier B1 is respectively connected with the input ends of a fourth resistor R4 and a sixth resistor R6, and the output end of the first operational amplifier B1 is respectively connected with the output end of a sixth resistor R6 and the input end of a second amplifier A2;
the output end of the fifth resistor R5 is connected with the input end of a first amplifier A1, the output end of a first amplifier A1 is connected with the positive input end of a second operational amplifier B2, the reverse input end of the second operational amplifier B2 is grounded GND, and the output end of a second operational amplifier B2 is respectively connected with the output end of a second amplifier A2, the base of a second NPN transistor Q2, the input end of an eighth resistor R8 and the input end of a second capacitor C2;
the output end of the recognizer is connected with the input end of a seventh resistor R7, the output end of a seventh resistor R7 is respectively connected with the collector of a second NPN transistor Q2 and the anode of a diode L, the emitter of the second NPN transistor Q2 is grounded GND, and the cathode of the diode L is respectively connected with the input ends of a tenth resistor R10 and a fifth capacitor C5;
an output end of the eighth resistor R8 is connected with a forward input end of a third operational amplifier B3, an inverting input end of the third operational amplifier B3 is connected with an input end of a third capacitor C3, an output end of the third operational amplifier B3 is connected with input ends of a fourth capacitor C4 and a ninth resistor R9 respectively, and output ends of the ninth resistor R9, a tenth resistor R10 and a fifth capacitor C5 are connected with collectors of a third NPN transistor Q3 and a fourth NPN transistor Q4 respectively;
a forward input end of the fourth operational amplifier B4 is connected to an output end of a third capacitor C3, a reverse input end of the fourth operational amplifier B4 is connected to an output end of a second capacitor C2, an output end of the fourth operational amplifier B4 is connected to an output end of a fourth capacitor C4 and a base of a third NPN transistor Q3, respectively, and an emitter of the third NPN transistor Q3 is connected to a base of the fourth NPN transistor Q4;
the input end of the first capacitor C1, the output end of the first resistor R1, the output end of the fourth resistor R4 and the emitter of the fourth NPN transistor Q4 are connected to ground, respectively.
The invention has the beneficial effects that:
the access control system based on the face recognition technology can cause insufficient image quality due to the change of environmental variables such as illumination brightness and the like, so that the recognition accuracy can be influenced. And noise can interfere the face recognition algorithm when the system converts actual influence into a digital image, so that the image processing algorithm needs to be deeply researched to ensure the image quality. The real-time problem of the system is always the focus of attention of the intelligent access control system, the system with good real-time performance can bring better user experience, and the problem that a machine is damaged by mistake due to the fact that a user is not patience caused by system delay can be avoided. Therefore, the system needs to select a proper image processing algorithm during design, and fully exerts the advantages of each processor of heterogeneous computing to improve the real-time performance of the system.
The algorithm is transplanted to an embedded platform based on ARM and FPGA architectures of a ZYNQ platform, the algorithm is conveniently improved and combined with an FPGA unit, and the identification accuracy and the real-time performance are met.
Drawings
FIG. 1 is a connection diagram of the laboratory access control verification system of the present invention;
FIG. 2 is a schematic diagram of a laboratory access control verification method of the present invention;
fig. 3 is a circuit diagram of an identification circuit of the laboratory access control verification system of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
The invention relates to an intelligent Internet of things access control system based on visible light, which combines a visible light acquisition technology, an Internet of things technology and an embedded type
The access control verification system with high safety, reliability and convenience is realized by applying the technology and the like, the access control and the background monitoring server are networked through the network, the centralized management and control of the access control system are realized, and the door is enabled to be based on the TCP/IP network protocol
The forbidden information exchange speed is higher, the supported number of the access control networking is larger, the networking can be carried out across regions or cities, and the remote access control system can realize remote access control,
And monitoring in real time across regions. The system can be generally divided into a face contour identification module; a facial feature recognition module; a face skin color identification module; and the identification degree comprehensive processing module. Each part is realized by adopting an independent design scheme according to different system functions and system requirements. Meanwhile, for the communication among all parts, the system pertinently uses a communication protocol and a data frame mode, is convenient for the system to carry out the communication of all parts, and simultaneously further enhances the reliability, the stability and the safety of the system.
The overall design idea of the system is as follows:
the face contour recognition module is used for recognizing the face contour data and confirming the identification degree LB of the face contour data;
the face facial features recognition module is used for recognizing the position and shape data of facial features and confirming the identification degree RB of the facial features data;
the face complexion identification module is used for identifying the complexion information of the face and confirming the face complexion data identification degree FB;
the identification degree comprehensive processing module is used for obtaining a laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB;
if the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
The most important thing of the access control system is to protect people's lives and properties reliably and safely. Compared with various access control systems widely existing in the current market, the system innovatively adopts 3 schemes to enhance the safety guarantee of the system in order to further improve the safety, and can be simply summarized into a mechanism for decomposing the face information into three types of data for comprehensive verification.
Further, the face contour recognition module recognizes the face contour data and determines the recognition degree LB of the face contour data, comprising the following steps:
(1.1) acquiring n characteristic points of all human face contours capable of being recognized, which are stored in the laboratory entrance guard verification system, and calculating the fluctuation level of each characteristic point of the human face contour;
Figure BDA0002545285010000111
wherein the content of the first and second substances,
Figure BDA0002545285010000121
as an intermediate parameter, Hbdi,jFor the jth feature point, σ, of the face contour of the ith person stored in the lab-gate verification systemjThe fluctuation level of the jth characteristic point of the face contour is m, the total amount of faces stored in the laboratory entrance guard verification system is i-1, 2, 3.. m, and j-1, 2,3 … n;
(1.2) calculating an adjusting coefficient of each feature point of the human face contour;
Figure BDA0002545285010000122
Hjadjusting coefficients of j characteristic points of the human face contour;
(1.3) acquiring n characteristic points of the face contour of the person to be recognized;
(1.4) calculating the difference between each characteristic point of the face contour of the person to be recognized and all the face contours of the recognizable persons stored in the laboratory entrance guard verification system;
Figure BDA0002545285010000123
Figure BDA0002545285010000124
the difference between the jth characteristic point of the face contour of the person to be recognized and the face contour of the jth characteristic point of the ith person stored in the laboratory entrance guard verification system is obtained;
(1.5) calculating the face contour data identification degree LB:
Figure BDA0002545285010000125
Figure BDA0002545285010000126
Figure BDA0002545285010000127
LB＝max(LBi,LB2,LB3,…,LBi…LBj)
wherein the content of the first and second substances,
Figure BDA0002545285010000128
LBiis an intermediate parameter;
and evaluating the conformity of the face contour curve of the identified person according to the size of the face contour data identification degree LB, wherein the smaller LB is, the lower the conformity of the identified person is.
The face contour is converted into the feature points of n face contours by utilizing the technology, so that a general concept of the face contour is converted into substantial data, LB can be obtained through calculation of the data, only a small number of feature points are extracted in the process, the calculated amount can be greatly reduced, meanwhile, when the adjustment coefficient of each feature point of the face contour is calculated in the process, different coefficients are given according to the change of the n feature points of the face contour of different persons in the laboratory forbidding door verification system, for the extracted feature points, the feature points at a certain position have low information content if all persons are different, smaller coefficients are given, and the feature points at different positions of different persons have large difference, then the information content of the feature points is large, giving a larger coefficient.
The facial features recognition module recognizes the position and shape data of facial features and confirms the recognition degree RB of the facial features data, and the facial features recognition module comprises the following steps:
firstly, acquiring the positions of five sense organs of a human face:
(2.1) carrying out convolution processing on the face image by using a 2D Gaussian filtering template and an original image to eliminate the noise of the face image;
(2.2) carrying out gray processing on the face image, and calculating the growth rate of pixel values along the horizontal left direction and the horizontal lower direction at each position of the gray image;
(2.3) calculating a comprehensive growth rate of each position;
Figure BDA0002545285010000131
wherein f (x, y) is the comprehensive growth rate of the position with the face image coordinate of (x, y), and HzIs the growth rate of the pixel value of the position with coordinates (x, y) of the face image along the horizontal left direction, HxThe growth rate of the pixel value of the position with the coordinate (x, y) of the face image along the horizontal downward direction;
(2.4) calculating a growth direction for each location;
Figure BDA0002545285010000132
wherein, the coordinate of the theta (x, y) face image is an included angle between the position of (x, y) along the horizontal left direction and the growth direction;
(2.5) traversing the whole face image, comparing whether the value of the pixel point at each position of the face image after gray scale is larger than the values of the pixel points at the front position and the rear position in the increasing direction of the position, if so, not changing the value of the pixel point at the position, otherwise, assigning the value of the pixel point at the position as 0;
(2.6) taking the point of the image processed in the step (2.5) with the value of the pixel point larger than a preset value T1 as a position point of the five sense organs, so as to obtain the coordinate positions of the five sense organs;
and then, taking the reciprocal of the Euclidean distance between the obtained coordinate position of the five sense organs and the preset coordinate position of the standard five sense organs as the data identification degree RB of the five sense organs of the face.
By utilizing the technology, the identification degree RB of the data of the five sense organs can be accurately obtained, in the process of obtaining the identification degree RB of the data of the five sense organs, firstly, the five sense organs of the human face are intelligently extracted, in the extraction process, the noise of the obtained human face image can be eliminated through (2.1), so that the interference information contained in the human face image for identifying the five sense organs at the back is greatly reduced, the human face image is easier to be intelligently identified, for the human face image at each position, and for the pixel point at each position of the human face image, the direction to which the pixel point at each position is changed to the greatest extent is found out by utilizing the increasing direction of the position, thereby the trend point of the five sense organs on the image, namely the outline of the five sense organs, can be found, meanwhile, the pixel point is judged to be not larger than the values before and after the increasing direction of the position, if the pixel value is larger than the value, therefore, the pixel values except for the five sense organs are removed, so that the pixel values of the five sense organs can be easily obtained, and the positions of the five sense organs of the human face can be automatically obtained.
The face complexion identification module is used for identifying complexion information of a face and confirming the face complexion data identification degree FB; the method comprises the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{Z，fs_min,fs_max}
wherein K is a skin color data set, Z is the brightness of the extracted face skin color, fs _ min is the minimum value of the pixels of the extracted face skin color, and fs _ max is the maximum value of the pixels of the extracted face skin color;
(3.2) obtaining the values of pixel points of the person to be identified, and calculating the weight of each pixel point value;
Figure BDA0002545285010000141
W_xsx,ythe weight value of the gray value of the pixel point of the position with the face image coordinate being (x, y),
H_xsx,ythe gray value of a pixel point of a position with the face image coordinate of (x, y);
h _ xs is the gray value of pixel points at all positions of the face image;
(3.3) calculating the belonged value of the skin color;
Figure BDA0002545285010000142
wherein G _ fsx,yThe probability of the skin color of the position with the coordinates of (x, y) of the face image is shown as cov, the covariance matrix after covariance of H _ xs by columns is shown as H _ xs,
Figure BDA0002545285010000143
to average the H xs column by column,
Figure BDA0002545285010000144
to pair
Figure BDA0002545285010000145
Transposition is carried out;
(3.4) confirming the identification degree FB of the face complexion data;
Figure BDA0002545285010000146
wherein, the (x, y) ∈ ν is the value of the coordinates (x, y) of the face image, and the value is all coordinate points of the face image.
The face complexion data identification FB can be easily obtained by utilizing the technology, in the process, firstly, the complexion belonging value is calculated according to the pixel point value, so that the probability that the certain pixel point value is used for determining complexion is judged, the probability that the certain pixel point value is used for determining the complexion probability value is judged, meanwhile, whether the complexion value is in the range of extracting the complexion of the face through the identification results of the face facial feature data identification RB and the face contour data identification LB is also judged, if the complexion value is not in the range, the pixel point of the position is used for determining whether the complexion probability is 0, and meanwhile, the face complexion data identification FB of the whole face image is determined through the weight of each pixel point value and the complexion belonging value.
Obtaining the laboratory entrance guard verification identification TB according to the following formula:
Figure BDA0002545285010000151
wherein, a and b are preset coefficients, the preset values are 1 +/-0.02, max (LB) is the maximum value that LB can take, max (RB) is the maximum value that RB can take, and max (FB) is the maximum value that FB can take.
If the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
Has the advantages that: the identification TB can be determined according to LB, RB and FB by the technology, so that the control method is used for controlling the release or the prohibition of the entrance guard of the laboratory.
Further, the face contour recognition module recognizes the face contour data and determines the recognition degree LB of the face contour data, including:
(1.1) inquiring the face outline data index X of the identified personmWherein m is 1,2,3, wherein XmFirst to third face contour models of a recognized person;
(1.2) calculating the face contour index adjustment coefficient R of the identified personi；
R1＝(X1-p1)/p1；R2＝(X2-p2)/p2；R3＝(X3-p3)/p3；
p·1Is X1The face contour ratio adjustment coefficient of (1); p is a radical of2Is X2The face contour ratio adjustment coefficient of (1); p is a radical of3Is X3The face contour ratio adjustment coefficient of (1);
(1.3) calculating a weighted face contour ratio adjustment coefficient P of the recognized person:
P＝1+R1X1+R2X2+R3X3；
(1.4) calling the average value V of the face contour model of the face contour of the identified person to obtain a value multiplier U of the face contour of the identified person,
U＝VP；
(1.5) calling all the collected face contour line data v of the identified personiI ∈ (1, 2, …, n) confirms the face contour data identification degree LB:
Figure BDA0002545285010000152
and evaluating the conformity of the face contour curve of the identified person through the size of the LB, wherein the smaller the LB, the lower the conformity of the identified person is.
The identification RB for confirming the facial feature data comprises the following steps:
(2.1) let the coordinates of the facial feature pixel iv be (X)iv，Yiv) Let the coordinates of g pixels of facial features not matching the coordinates be (X)1，Y1)，(X2，Y2)…(Xg，Yg)，g≥3；
(2.2) the Euclidean distance between the facial five sense organ pixels which do not coincide with the coordinates and the associated m facial five sense organ pixels of the person is as follows:
Figure BDA0002545285010000153
the above equation is derived:
Figure BDA0002545285010000154
the formula is calculated through a simplified structure to obtain the adjustment coordinates of facial feature pixels which are not matched with the coordinates:
Figure BDA0002545285010000161
Figure BDA0002545285010000162
Figure BDA0002545285010000163
Figure BDA0002545285010000164
(2.3) setting the deviation distance r of the facial feature pixel iu and a standard storage facial feature pixel model, wherein the communication error of a data acquisition image system is r0(ii) a The average power value of the face acquisition system is P (r)0) (ii) a The image receiving power of the facial five-organ pixels which are not matched with the coordinates is P (r); then:
Figure BDA0002545285010000165
alpha is an image acquisition path loss index, and the deviation distance r between the facial feature pixel iu and a standard stored facial feature pixel model is obtained through calculation;
(2.4) calculating and confirming the identification degree RB of the facial feature data;
Figure BDA0002545285010000166
d0is the detection distance, X, of the system from the acquired faceσIs a positive distribution of random values with a standard deviation of [3, 6 ]]A (i) data acquisition loss rate for facial features; h0The original signal strength of the original facial features data is collected for the system.
The confirmation of the identification degree FB of the face skin color data comprises the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{k0,k1,k2,…,kv}；
kvstoring data for a skin tone database; k is a radical of0As standard skin tone data, k0The deviation data of (a) includes:
me(k0)＝{k01,k02,k02,…,k0t}；
t is the number of deviation data;
(3.2) estimating K and measured skin tone data sqThe color difference between the two images is obtained,
Figure BDA0002545285010000167
Rele(W,G(me(k0) ) represents the correlation degree of the skin color data corresponding to the face skin color database and the measured skin color data, G(s)q) A set of correlation values representing measured skin color data obtained by the acquisition;
(3.3) confirming the identification degree FB of the face complexion data
Figure BDA0002545285010000168
Wei denotes K and CSCorrelation between, D (K, C)S) Represents K and CSThe color difference value between them.
The obtained laboratory entrance guard verification identification TB is as follows:
Figure BDA0002545285010000171
body identification coefficient P ═ alpha (LB + RB)t
Weak identification coefficient AOC is 24.2 XFB2。
The invention also relates to a laboratory entrance guard verification method, which is characterized by comprising the following steps:
(1) recognizing the face contour data and confirming the recognition degree LB of the face contour data;
(1.1) inquiring the face outline data index X of the identified personmWherein m is 1,2,3, wherein XmFirst to third face contour models of a recognized person;
(1.2) calculating the face contour index adjustment coefficient R of the identified personi；
R1＝(X1-p1)/p1；R2＝(X2-p2)/p2；R3＝(X3-p3)/p3；
p·1Is X1The face contour ratio adjustment coefficient of (1); p is a radical of2Is X2The face contour ratio adjustment coefficient of (1); p is a radical of3Is X3The face contour ratio adjustment coefficient of (1);
(1.3) calculating a weighted face contour ratio adjustment coefficient P of the recognized person:
P＝1+R1X1+R2X2+R3X3；
(1.4) calling the average value V of the face contour model of the face contour of the identified person to obtain a value multiplier U of the face contour of the identified person,
U＝VP；
(1.5) calling all the collected face contour line data v of the identified personiI ∈ (1, 2, …, n) confirms the face contour data identification degree LB:
Figure BDA0002545285010000172
evaluating the conformity of the human face contour curve of the identified person according to the size of LB, wherein the smaller LB is, the lower the conformity of the identified person is;
(2) recognizing the position and shape data of the facial features, and confirming the identification RB of the facial features;
(2.1) let the coordinates of the facial feature pixel iv be (X)iv，Yiv) Let the coordinates of g pixels of facial features not matching the coordinates be (X)1，Y1)，(X2，Y2)…(Xg，Yg)，g≥3；
(2.2) the Euclidean distance between the facial five sense organ pixels which do not coincide with the coordinates and the associated m facial five sense organ pixels of the person is as follows:
Figure BDA0002545285010000173
the above equation is derived:
Figure BDA0002545285010000174
the formula is calculated through a simplified structure to obtain the adjustment coordinates of facial feature pixels which are not matched with the coordinates:
Figure BDA0002545285010000175
Figure BDA0002545285010000176
Figure BDA0002545285010000181
Figure BDA0002545285010000182
(2.3) setting the deviation distance r of the facial feature pixel iu and a standard storage facial feature pixel model, wherein the communication error of a data acquisition image system is r0(ii) a The average power value of the face acquisition system is P (r)0) (ii) a The image receiving power of the facial five-organ pixels which are not matched with the coordinates is P (r); then:
Figure BDA0002545285010000183
alpha is an image acquisition path loss index, and the deviation distance r between the facial feature pixel iu and a standard stored facial feature pixel model is obtained through calculation;
(2.4) calculating and confirming the identification degree RB of the facial feature data;
Figure BDA0002545285010000184
d0is the detection distance, X, of the system from the acquired faceσIs a positive distribution of random values with a standard deviation of [3, 6 ]]A (i) data acquisition loss rate for facial features; h0Collecting the original signal intensity of original facial feature data for a system;
(3) the face complexion identification module is used for identifying the complexion information of the face and confirming the face complexion data identification degree FB;
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{k0,k1,k2,…,kv}；
kvstoring data for a skin tone database; k is a radical of0As standard skin tone data, k0The deviation data of (a) includes:
me(k0)＝{k01,k02,k02,…,k0t}；
t is the number of deviation data;
(3.2) estimating K and measured skin tone data sqThe color difference between the two images is obtained,
Figure BDA0002545285010000185
Rele(W,G(me(k0) ) represents the correlation degree of the skin color data corresponding to the face skin color database and the measured skin color data, G(s)q) A set of correlation values representing measured skin color data obtained by the acquisition;
(3.3) confirming the identification degree FB of the face complexion data
Figure BDA0002545285010000186
Wei denotes K and CSCorrelation between, D (K, C)S) Represents K and CSThe color difference value between;
the identification degree comprehensive processing module is used for obtaining a laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB;
if the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
The obtained laboratory entrance guard verification identification TB is as follows:
Figure BDA0002545285010000191
body identification coefficient P ═ alpha (LB + RB)t
Weak identification coefficient AOC is 24.2 XFB2。
According to the method, three different information data of the face are collected and processed, a comprehensive verification value is obtained through a coefficient relation, the accuracy of the information extraction process and the accuracy of the verification process are higher, and due to the change of environmental variables such as the illumination brightness degree and other reasons, the access control system based on the face recognition technology can cause insufficient image quality, so that the recognition accuracy can be influenced. And noise can interfere the face recognition algorithm when the system converts actual influence into a digital image, so that the image processing algorithm needs to be deeply researched to ensure the image quality. The real-time problem of the system is always the focus of attention of the intelligent access control system, the system with good real-time performance can bring better user experience, and the problem that a machine is damaged by mistake due to the fact that a user is not patience caused by system delay can be avoided. Therefore, the system needs to select a proper image processing algorithm during design, and fully exerts the advantages of each processor of heterogeneous computing to improve the real-time performance of the system. The algorithm is transplanted to an embedded platform based on ARM and FPGA architectures of a ZYNQ platform, the algorithm is conveniently improved and combined with an FPGA unit, and the identification accuracy and the real-time performance are met.
The embodiment of the invention provides a laboratory entrance guard verification system, wherein a face contour recognition module comprises a recognition circuit, as shown in fig. 3, the recognition circuit comprises: a first resistor R1, a second resistor R2, a third resistor R3, a fourth resistor R4, a fifth resistor R5, a sixth resistor R6, a seventh resistor R7, an eighth resistor R8, a ninth resistor R9, a tenth resistor R10, a first capacitor C1, a second capacitor C2, a third capacitor C3, a fourth capacitor C4, a fifth capacitor C5, a first NPN transistor Q1, a second NPN transistor Q2, a third NPN transistor Q3, a fourth NPN transistor Q4, a diode L, a first operational amplifier B1, a second operational amplifier B2, a third operational amplifier B3, a fourth operational amplifier B4, a first amplifier a1, a second amplifier a2, a discriminator, a power source VCC, and a ground GND;
the power supply VCC is respectively connected with the input ends of the identifier, the second resistor R2, the fifth resistor R5 and the first capacitor C1, the output end of the second resistor R2 is respectively connected with the input end of the third resistor R3 and the collector of the first NPN transistor Q1, the output end of the first capacitor C1 is connected with the base of the first NPN transistor Q1, and the emitter of the first NPN transistor Q1 is connected with the input end of the first resistor R1;
the output end of the third resistor R3 is connected with the positive input end of a first operational amplifier B1, the inverting input end of the first operational amplifier B1 is respectively connected with the input ends of a fourth resistor R4 and a sixth resistor R6, and the output end of the first operational amplifier B1 is respectively connected with the output end of a sixth resistor R6 and the input end of a second amplifier A2;
the output end of the fifth resistor R5 is connected with the input end of a first amplifier A1, the output end of a first amplifier A1 is connected with the positive input end of a second operational amplifier B2, the reverse input end of the second operational amplifier B2 is grounded GND, and the output end of a second operational amplifier B2 is respectively connected with the output end of a second amplifier A2, the base of a second NPN transistor Q2, the input end of an eighth resistor R8 and the input end of a second capacitor C2;
the output end of the recognizer is connected with the input end of a seventh resistor R7, the output end of a seventh resistor R7 is respectively connected with the collector of a second NPN transistor Q2 and the anode of a diode L, the emitter of the second NPN transistor Q2 is grounded GND, and the cathode of the diode L is respectively connected with the input ends of a tenth resistor R10 and a fifth capacitor C5;
an output end of the eighth resistor R8 is connected with a forward input end of a third operational amplifier B3, an inverting input end of the third operational amplifier B3 is connected with an input end of a third capacitor C3, an output end of the third operational amplifier B3 is connected with input ends of a fourth capacitor C4 and a ninth resistor R9 respectively, and output ends of the ninth resistor R9, a tenth resistor R10 and a fifth capacitor C5 are connected with collectors of a third NPN transistor Q3 and a fourth NPN transistor Q4 respectively;
a forward input end of the fourth operational amplifier B4 is connected to an output end of a third capacitor C3, a reverse input end of the fourth operational amplifier B4 is connected to an output end of a second capacitor C2, an output end of the fourth operational amplifier B4 is connected to an output end of a fourth capacitor C4 and a base of a third NPN transistor Q3, respectively, and an emitter of the third NPN transistor Q3 is connected to a base of the fourth NPN transistor Q4;
the input end of the first capacitor C1, the output end of the first resistor R1, the output end of the fourth resistor R4 and the emitter of the fourth NPN transistor Q4 are connected to ground, respectively.
The beneficial effects of the above technical scheme are: by arranging the third NPN transistor Q3 and the fourth NPN transistor Q4, abnormal signals in the identification circuit can be filtered, so that the voltages of the third NPN transistor and the fourth NPN transistor are conducted, and a protection effect is achieved;
through setting up first operational amplifier B1, second operational amplifier B2, first amplifier A1, second amplifier A2, improved identification circuit to external interference killing feature, and improved this identification circuit's stability, through setting up the recognizer, can effectually discern the human face.
The above description is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (3)

1. A laboratory access verification system, comprising:
the face contour recognition module is used for recognizing the face contour data and confirming the identification degree LB of the face contour data;
the face facial features recognition module is used for recognizing the position and shape data of facial features and confirming the identification degree RB of the facial features data;
the face complexion identification module is used for identifying the complexion information of the face and confirming the face complexion data identification degree FB;
the identification degree comprehensive processing module is used for obtaining a laboratory entrance guard verification identification degree TB according to the face contour data identification degree LB, the face facial feature data identification degree RB and the face skin color data identification degree FB;
if the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; if the laboratory entrance guard verification identification TB is less than the threshold value T, the entrance guard is kept forbidden;
the face contour recognition module recognizes face contour data and confirms the face contour data identification degree LB, and comprises the following steps:
(1.1) acquiring n characteristic points of all human face contours capable of being recognized, which are stored in the laboratory entrance guard verification system, and calculating the fluctuation level of each characteristic point of the human face contour;
Figure FDA0002545283000000011
wherein the content of the first and second substances,
Figure FDA0002545283000000012
as an intermediate parameter, Hbdi,jFor the jth feature point, σ, of the face contour of the ith person stored in the lab-gate verification systemjThe fluctuation level of the jth characteristic point of the face contour is m, the total amount of faces stored in the laboratory entrance guard verification system is i-1, 2, 3.. m, and j-1, 2,3 … n;
(1.2) calculating an adjusting coefficient of each feature point of the human face contour;
Figure FDA0002545283000000013
Hjadjusting coefficients of j characteristic points of the human face contour;
(1.3) acquiring n characteristic points of the face contour of the person to be recognized;
(1.4) calculating the difference between each characteristic point of the face contour of the person to be recognized and all the face contours of the recognizable persons stored in the laboratory entrance guard verification system;
Figure FDA0002545283000000014
Figure FDA0002545283000000015
the difference between the jth characteristic point of the face contour of the person to be recognized and the face contour of the jth characteristic point of the ith person stored in the laboratory entrance guard verification system is obtained;
(1.5) calculating the face contour data identification degree LB:
Figure FDA0002545283000000021
Figure FDA0002545283000000022
Figure FDA0002545283000000023
LB＝max(LBi,LB2,LB3,…,LBi…LBj)
wherein the content of the first and second substances,
Figure FDA0002545283000000024
LBiis an intermediate parameter;
and evaluating the conformity of the face contour curve of the identified person according to the size of the face contour data identification degree LB, wherein the smaller LB is, the lower the conformity of the identified person is.
2. The laboratory entrance guard verification system according to claim 1, wherein said face skin color identification module identifies skin color information of a face and confirms a face skin color data identification FB; the method comprises the following steps:
(3.1) extracting skin color data corresponding to the face skin color database according to the identification results of the identification degree RB of the facial feature data and the identification degree LB of the facial contour data:
K＝{Z，fs_min,fs_max}
wherein K is a skin color data set, Z is the brightness of the extracted face skin color, fs _ min is the minimum value of the pixels of the extracted face skin color, and fs _ max is the maximum value of the pixels of the extracted face skin color;
(3.2) obtaining the values of pixel points of the person to be identified, and calculating the weight of each pixel point value;
Figure FDA0002545283000000025
W_xsx,ythe weight value of the gray value of the pixel point of the position with the face image coordinate being (x, y),
H_xsx,ythe gray value of a pixel point of a position with the face image coordinate of (x, y);
h _ xs is the gray value of pixel points at all positions of the face image;
(3.3) calculating the belonged value of the skin color;
Figure FDA0002545283000000026
wherein G _ fsx,yThe probability of the skin color of the position with the coordinates of (x, y) of the face image is shown as cov, the covariance matrix after covariance of H _ xs by columns is shown as H _ xs,
Figure FDA0002545283000000031
to average the H xs column by column,
Figure FDA0002545283000000032
to pair
Figure FDA0002545283000000033
Transposition is carried out;
(3.4) confirming the identification degree FB of the face complexion data;
Figure FDA0002545283000000034
wherein, the (x, y) ∈ ν is the value of the coordinates (x, y) of the face image, and the value is all coordinate points of the face image.
3. The laboratory entrance guard verification system of claim 1, wherein the laboratory entrance guard verification identification TB is obtained according to the following formula:
Figure FDA0002545283000000035
wherein, a and b are preset coefficients, the preset values are 1 +/-0.02, max (LB) is the maximum value that LB can take, max (RB) is the maximum value that RB can take, and max (FB) is the maximum value that FB can take.
If the laboratory entrance guard verification identification TB is larger than or equal to the threshold T, the entrance guard is forbidden; and if the laboratory entrance guard verification identification TB is less than the threshold T, the entrance guard is kept forbidden.
CN202010558650.3A 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method Active CN111932754B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010558650.3A CN111932754B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910764625.8A CN110555929B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method
CN202010558650.3A CN111932754B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
CN201910764625.8A Division CN110555929B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method

Publications (2)

Publication Number Publication Date
CN111932754A true CN111932754A (en) 2020-11-13
CN111932754B CN111932754B (en) 2021-12-28

Family

ID=68737699

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201910764625.8A Active CN110555929B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method
CN202010558650.3A Active CN111932754B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN201910764625.8A Active CN110555929B (en) 2019-08-19 2019-08-19 Laboratory entrance guard verification system and verification method

Country Status (1)

Country Link
CN (2) CN110555929B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114005160A (en) * 2021-10-28 2022-02-01 建湖县公安局 Access control system and method based on identity two-dimensional code and artificial intelligence

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI787664B (en) * 2020-10-23 2022-12-21 正修學校財團法人正修科技大學 IoT SECURITY SYSTEM AND METHOD THEREOF

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102271241A (en) * 2011-09-02 2011-12-07 北京邮电大学 Image communication method and system based on facial expression/action recognition
CN105023010A (en) * 2015-08-17 2015-11-04 中国科学院半导体研究所 Face living body detection method and system
CN105374098A (en) * 2015-12-14 2016-03-02 天津光电通信技术有限公司 Method used for unlocking using human body double-characteristic identification module
CN105427421A (en) * 2015-11-16 2016-03-23 苏州市公安局虎丘分局 Entrance guard control method based on face recognition
CN107317846A (en) * 2017-06-07 2017-11-03 上海储翔信息科技有限公司 Vehicle management system based on cloud platform
CN107992797A (en) * 2017-11-02 2018-05-04 中控智慧科技股份有限公司 Face identification method and relevant apparatus
CN109636397A (en) * 2018-11-13 2019-04-16 平安科技（深圳）有限公司 Transit trip control method, device, computer equipment and storage medium
CN109727350A (en) * 2018-12-14 2019-05-07 深圳壹账通智能科技有限公司 A kind of Door-access control method and device based on recognition of face
CN109767531A (en) * 2018-12-29 2019-05-17 三石量子（苏州）信息科技有限公司 A kind of Intelligent human-face identification gate control system

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10558848B2 (en) * 2017-10-05 2020-02-11 Duelight Llc System, method, and computer program for capturing an image with correct skin tone exposure
CN106709411A (en) * 2015-11-17 2017-05-24 腾讯科技（深圳）有限公司 Appearance level acquisition method and device
CN106897590A (en) * 2015-12-17 2017-06-27 阿里巴巴集团控股有限公司 The method of calibration and device of figure information
CN106407909B (en) * 2016-08-31 2019-04-02 北京贝塔科技股份有限公司 Face recognition method, device and system
CN106778468B (en) * 2016-11-14 2019-08-30 深圳奥比中光科技有限公司 3D face identification method and equipment
CN207337495U (en) * 2017-07-11 2018-05-08 常熟理工学院 Laboratory Intelligent entrance guard management system based on recognition of face
CN109544754B (en) * 2018-11-27 2021-09-28 湖南视觉伟业智能科技有限公司 Access control method and system based on face recognition and computer equipment
CN110084925A (en) * 2019-04-03 2019-08-02 南京邮电大学 A kind of university dormitory management method and its system based on recognition of face

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102271241A (en) * 2011-09-02 2011-12-07 北京邮电大学 Image communication method and system based on facial expression/action recognition
CN105023010A (en) * 2015-08-17 2015-11-04 中国科学院半导体研究所 Face living body detection method and system
CN105427421A (en) * 2015-11-16 2016-03-23 苏州市公安局虎丘分局 Entrance guard control method based on face recognition
CN105374098A (en) * 2015-12-14 2016-03-02 天津光电通信技术有限公司 Method used for unlocking using human body double-characteristic identification module
CN107317846A (en) * 2017-06-07 2017-11-03 上海储翔信息科技有限公司 Vehicle management system based on cloud platform
CN107992797A (en) * 2017-11-02 2018-05-04 中控智慧科技股份有限公司 Face identification method and relevant apparatus
CN109636397A (en) * 2018-11-13 2019-04-16 平安科技（深圳）有限公司 Transit trip control method, device, computer equipment and storage medium
CN109727350A (en) * 2018-12-14 2019-05-07 深圳壹账通智能科技有限公司 A kind of Door-access control method and device based on recognition of face
CN109767531A (en) * 2018-12-29 2019-05-17 三石量子（苏州）信息科技有限公司 A kind of Intelligent human-face identification gate control system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114005160A (en) * 2021-10-28 2022-02-01 建湖县公安局 Access control system and method based on identity two-dimensional code and artificial intelligence

Also Published As

Publication number Publication date
CN110555929A (en) 2019-12-10
CN110555929B (en) 2020-08-14
CN111932754B (en) 2021-12-28

Similar Documents

Publication Publication Date Title
CN108921100B (en) Face recognition method and system based on visible light image and infrared image fusion
Sun et al. Improving iris recognition accuracy via cascaded classifiers
JP3469031B2 (en) Face image registration apparatus and method
CN100414558C (en) Automatic fingerprint distinguishing system and method based on template learning
CN109882019B (en) Automobile electric tail door opening method based on target detection and motion recognition
CN110728225B (en) High-speed face searching method for attendance checking
CN109766836B (en) Finger vein identification unlocking method and system
CN106372611A (en) Identity identification method and system based on palm vein features, and image acquisition apparatus
CN105427421A (en) Entrance guard control method based on face recognition
CN107392112A (en) A kind of facial expression recognizing method and its intelligent lock system of application
CN201845343U (en) Image acquisition device for human face recognition
CN111932754B (en) Laboratory entrance guard verification system and verification method
CN106250825A (en) A kind of at the medical insurance adaptive face identification system of applications fields scape
CN108446687B (en) Self-adaptive face vision authentication method based on interconnection of mobile terminal and background
CN102900305A (en) Keyless safe box system based on human face recognition
CN113449704B (en) Face recognition model training method and device, electronic equipment and storage medium
CN109377601B (en) Intelligent office access control system based on fingerprint identification
CN205644823U (en) Social security self -service terminal device
CN111062021A (en) Method and device for identity authentication based on wearable equipment
CN106355150A (en) Universal fingerprint recognition system and method
CN108288040A (en) Multi-parameter face identification system based on face contour
CN106447840A (en) Multifunctional intelligent entrance guard system
CN206741505U (en) Identify management equipment and device
Amjed et al. Noncircular iris segmentation based on weighted adaptive hough transform using smartphone database
CN111932758A (en) Intelligent access control system based on face recognition technology

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant