CN111475791A - High-security face recognition method, verification terminal and storage medium - Google Patents
High-security face recognition method, verification terminal and storage medium Download PDFInfo
- Publication number
- CN111475791A CN111475791A CN202010285942.4A CN202010285942A CN111475791A CN 111475791 A CN111475791 A CN 111475791A CN 202010285942 A CN202010285942 A CN 202010285942A CN 111475791 A CN111475791 A CN 111475791A
- Authority
- CN
- China
- Prior art keywords
- verification
- user
- eyeball
- verification terminal
- terminal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
- G06F3/013—Eye tracking input arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- 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
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- 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
- G06V40/18—Eye characteristics, e.g. of the iris
Abstract
The invention provides a high-security face recognition method, a verification terminal and a storage medium, wherein the method comprises the following steps: the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule; the verification terminal shoots a face image and an eyeball motion track of a user, judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not, and traverses a cloud database to judge whether a face verification image matched with the shot face image exists or not; and when the similarity between the eyeball motion track and the verification line reaches a similarity threshold value and a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed. The high safety of the face recognition is ensured by combining the double-layer verification of the face recognition and the eyeball motion discrimination.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a high-security face recognition method, a verification terminal and a storage medium.
Background
The face recognition is a biological feature recognition technology for identity authentication based on human physiognomic feature information, adopts a non-contact mode for recognition, and can quickly and sanitarily pass through the identity authentication, so the face recognition technology is widely applied to common application fields such as the entering of a high-speed rail station through face recognition, the attendance verification by face recognition, the face recognition payment during supermarket self-service payment, the opening of a cabinet door by face recognition during express taking of an intelligent cabinet and the like.
Along with the popularization of face recognition technology, more and more face recognition cracking technologies are also developed, for example, a 3D printing technology for copying face features; a face-lifting technology developed at a high speed is to change the characteristics of human faces; the method for stealing the face features comprises the following steps: the face features are collected through some mobile phone applications, or personal face video information shot on the mobile phone is stolen by the outside, or objects such as a hidden camera, a mirror and the like are arranged in some occasions, so that the personal face feature information and the like are obtained secretly. Once the face information of an individual is illegally stolen, the stolen face information is illegally used by a stealer, and reputation damage or economic loss is certainly brought to the stealer.
The face recognition cracking technology is generated just because the safety level of the face recognition technology at the present stage is not high, when the face recognition technology is applied to the verification terminal for face recognition, a thief can easily pass through the face recognition verification of the verification terminal, and then the identity of the thief is used for carrying out certain economic transactions or entering specific occasions, so that reputation damage or economic loss is brought to the thief.
Therefore, the prior art has defects and needs to be improved and developed.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a high-security face recognition method, a verification terminal and a storage medium, aiming at solving the problem of low security level of the face recognition technology in the prior art.
The technical scheme adopted by the invention for solving the technical problem is as follows:
a high-density face recognition method comprises the following steps:
the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule;
the verification terminal shoots a face image and an eyeball motion track of a user, judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not, and traverses a cloud database to judge whether a face verification image matched with the shot face image exists or not;
and when the similarity between the eyeball motion track and the verification line reaches a similarity threshold value and a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed.
Further, the method for judging whether the similarity between the eyeball motion trajectory and the verification line reaches a similarity threshold value includes the steps of:
the verification terminal shoots pictures of eyeballs of the user according to a time axis and connects all eyeball focuses in the shot pictures to form an eyeball motion track;
and the verification terminal compares the eyeball motion track with the verification line and judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not.
Further, the verification terminal shoots the picture of the eyeballs of the user according to a time axis, all the eyeballs in the shot picture are connected to form an eyeball motion track, and the method also comprises the following steps:
the verification terminal judges whether the areas of the irises and the sclera in the eyeballs corresponding to all the focuses on the eyeball motion track formed by connecting the verification terminal with the focuses are close to the areas of the irises and the sclera in the shot picture or not according to the shot picture;
if the two points are close to each other, the verification terminal compares the eyeball motion track formed by connecting all the focuses with the verification line;
if the positions of the iris and the sclera in the shot picture are not similar, the verification terminal fits a new eyeball motion track according to the areas of the iris and the sclera in the shot picture and the position of the focus on the eyeball motion track, and the new eyeball motion track is compared with the verification line.
Furthermore, the verification terminal is provided with two cameras for respectively shooting eyeballs of two eyes of the user, obtains a focal length between the two eyes by analyzing the positions of the focal points of the shot eyeballs of the two eyes, and assists the formation of the eyeball movement track according to the focal length between the two eyes.
Further, the formation of the eyeball movement trajectory is assisted by the verification terminal according to the focal length between the two eyes, and the method specifically includes:
the movement of the eyeballs of the two eyes of the user corresponds to an eyeball movement track respectively, and the eyeball movement tracks of the two eyes are symmetrical by taking the focal length of the two eyes as a boundary;
when the eyeball motion tracks of the two eyes are asymmetric, the verification terminal combines the focal length and the focal point of the two eyes and the area of the iris and the sclera in the shot eyeball picture, and adjusts the other eyeball motion track by taking one eyeball motion track as a reference according to a similar principle.
Further, the way that the verification terminal generates the verification line according to the preset rule is specifically that:
the verification terminal calculates the product of the current date and the month according to the current date and the month;
when the verification terminal receives an identity verification instruction of a to-be-verified person, the verification terminal acquires the current hour time, and the hour time is increased on the basis of the product of the current date and the month to form a verification line and store the verification line in the verification terminal.
Further, the way to calculate the similarity threshold is:
and the verification terminal performs numerical analysis on the eyeball motion track, and calculates the numerical difference between the verification line and the eyeball motion track to obtain the similarity threshold.
Further, a verification terminal shoots a face image and an eyeball motion track of a user, and whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not is judged, wherein the mode of inputting the eyeball motion track of the user by the verification terminal comprises the following steps:
the verification terminal receives an input starting instruction sent by a user, captures a focus of eyeballs of the user according to the input starting instruction, carries out focusing processing, and prompts the user to send a first input instruction after focusing is finished;
the verification terminal shoots a first eyeball movement track generated by the first rotation of the eyeballs of the user according to the first input instruction, and prompts the user to send a second input instruction;
the verification terminal shoots a second eyeball movement track generated by the second rotation of the eyeballs of the user according to the second input instruction, and prompts the user to send a third input instruction;
the steps are repeated until the verification terminal shoots the last eyeball movement track generated by the last rotation of the eyeballs of the user;
and the verification terminal locks the end point of the last eyeball movement track to finish the input of all the eyeball movement tracks of the user.
Further, the verification terminal shoots the face image and the eyeball motion trail of the user, and simultaneously the verification terminal further comprises:
verifying the visual distance of the terminal when positioning the eyeball of the user, and judging whether the change of the positioned visual distance is matched with the movement track of the eyeball of the user;
if the eyeball is matched with the verification line, the eyeball of the user is real, and the verification terminal judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not;
if not, the user authentication fails, and the authentication is quitted.
The invention also discloses a verification terminal, which comprises a processor and a memory connected with the processor, wherein the memory stores a configuration program of the high-security face recognition method, and the configuration program of the high-security face recognition method realizes the high-security face recognition method when being executed by the processor.
The invention also discloses a storage medium, wherein the storage medium stores a computer program which can be executed for realizing the high-security face recognition method.
The invention provides a high-security face recognition method, a verification terminal and a storage medium, wherein the method comprises the following steps: the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule; the verification terminal shoots a face image and an eyeball motion track of a user, judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not, and traverses a cloud database to judge whether a face verification image matched with the shot face image exists or not; and when the similarity between the eyeball motion track and the verification line reaches a similarity threshold value and a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed. The verification terminal generates a verification line according to a preset rule, the user rotates an eyeball according to the preset rule, the verification terminal shoots an eyeball movement track of the user, the face characteristics of the user are obtained, the user passes the identity verification under the condition that the face characteristics and the eyeball movement track are identified, the real face can be judged by identifying the face through double-layer verification, the safety level of face identification performed by the verification terminal is improved, and the high safety and accuracy of face identification are ensured.
Drawings
FIG. 1 is a flow chart of a high-security face recognition method according to a preferred embodiment of the present invention;
FIG. 2 is a flowchart illustrating a preferred embodiment of the verification terminal determining similarity between eye movement trajectory and verification line according to the present invention;
FIG. 3 is a flowchart illustrating an embodiment of the present invention for verifying the input of the eye movement trace by the terminal;
FIG. 4 is a flow chart of a preferred embodiment of the present invention for verifying the identity of a user by a verification terminal after adding a Z-axis advanced stealth algorithm;
fig. 5 is a functional block diagram of a preferred embodiment of the authentication terminal of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer and clearer, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1, fig. 1 is a flowchart of a high-security face recognition method according to the present invention. As shown in fig. 1, a high-security face recognition method according to an embodiment of the present invention includes the following steps:
s100, the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule.
Specifically, when the user performs identity authentication on the authentication terminal, the authentication function of the authentication terminal is started, the authentication terminal automatically generates an authentication line according to an algorithm set in the system so as to perform eyeball authenticity authentication on the user, wherein the preset rule is designed by the user in advance and set in the authentication terminal, and therefore the user knows the preset rule.
The verification terminal is widely applied, can be applied to cash registers, attendance machines, gate machines, mobile phone opening authentication, APP login use, intelligent express cabinets, library book borrowing and the like, can be realized through the technical scheme of the invention as long as a face recognition scene can be used, and is beneficial to improving the safety level of face authentication.
There are various schemes for generating the verification line according to the preset rule, which are exemplified here:
in the first scheme, the preset rule is fixed, for example, a fixed password, and an authentication line formed by the fixed password is not changed in a certain period of time, and when the authentication line needs to be changed, the rule needs to be reset. The fixed password can be a combination of 26 letters or any combination of Arabic numerals 1-10.
And in a second scheme, the preset rule is regularly circulated, for example, the preset rule is set as the current time, and the current time can be selected, so that the user can automatically rotate the eyeball by knowing the current time, and the authentication terminal can conveniently perform identity authentication on the user.
In a preferred embodiment, in order to enhance the unpredictability of the verification line, the verification terminal in the invention obtains the verification line according to the product of the current date and the month and then the time of the current hour hand.
Since there are various ways to generate the verification line, it is not described herein again. It can be understood that, as long as the verification line can be generated in the verification terminal, and the difficulty level of the verification line can be adjusted according to a specific application scenario, all the ways having the verification line characteristics described in the present invention belong to the protection scope of the present invention.
S200, shooting a facial image and an eyeball motion track of a user by a verification terminal, and judging whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not; and traversing the cloud database to judge whether a face verification image matched with the shot face image exists.
Specifically, the verification terminal is provided with two cameras for shooting eyeballs of the user, and the face of the user shot by the cameras comprises face features and an eyeball movement track of the user.
The cloud database is associated with the verification terminal, and as long as personal information (at least including name and face characteristics and identity numbers, company names, personal titles and the like in different occasions) of a user is stored in the cloud database in advance, all verification terminals which can be accessed into the cloud database can search face verification images in the cloud database, so that occupation of an internal memory of the verification terminal is avoided, the use range of the verification terminal is expanded, and one cloud database can provide data services for a plurality of verification terminals at the same time.
In the embodiment of the present invention, the verification terminal shoots the eye movement trajectory of the user, and determines whether the similarity between the eye movement trajectory and the verification line reaches the similarity threshold, as shown in fig. 2, fig. 2 is a flowchart of a preferred embodiment of the verification terminal for determining the similarity between the eye movement trajectory and the verification line in the present invention, which specifically includes the following steps:
s210, the verification terminal shoots the picture of the eyeballs of the user according to a time axis, and all the eyeballs in the shot picture are connected to form an eyeball motion track.
Specifically, the verification terminal tracks the focus of the eyeball through the camera, records the position of the focus when the position of the focus changes, and finally connects all recorded focuses to form an eyeball motion track.
Meanwhile, the verification terminal can automatically analyze the face picture shot by the camera to judge whether the eyeball motion track analyzed in the verification terminal is consistent with the actual eyeball rotation track of the user or not, so that the eyeball motion track analyzed by the verification terminal is real and effective, and the safety of face verification is improved. Specifically, step S220: and the verification terminal analyzes whether the change of the area of the iris and the sclera generated by the eyeball motion track conforms to the rule or not according to the change of the area of the iris and the sclera in the photographed eyeball.
The step S220 specifically includes:
s221, the verification terminal judges whether the area of the iris and the sclera in the eyeball corresponding to all the focuses on the eyeball motion track formed by connecting the verification terminal with the focuses is close to the area of the iris and the sclera in the shot picture or not according to the shot picture.
Specifically, the eyeball is constructed by an iris, a pupil and a sclera, wherein the pupil and the iris are black, namely, eyeball, and the sclera is white, namely, eye white.
And S222, if the two are close, comparing the eyeball motion tracks formed by connecting all the focuses with the verification lines by the verification terminal. Specifically, if the area change of the eyeball and the white of the eye caused by the actual rotation of the eyeball is similar to the area change of the eyeball and the white of the eye corresponding to the focal position captured by the verification terminal, it is indicated that the focal point judged by the verification terminal through machine vision is similar to the actual situation in height, and the verification terminal analyzes the eyeball motion track formed by all the focal points through the parallax principle.
And S223, if the areas of the iris and the sclera in the shot picture are not similar, the verification terminal fits a new eyeball motion track according to the areas of the iris and the sclera in the shot picture and the position of the focus on the eyeball motion track, and the new eyeball motion track is compared with the verification line. Specifically, the verification terminal compares the area ratio of the iris and the sclera in the picture with the focal position captured by the verification terminal, and under the condition that the area ratio of the iris and the sclera is known, the position of the focal point is appropriately moved, so that the position of the focal point corresponds to the area ratio of the iris and the sclera. It can be understood that the position of the moving focus is moved in the surrounding position range based on the captured focus position, and the new eyeball motion trajectory formed by fitting is the connecting line of all the focuses after the focus position is moved.
In another embodiment of the present invention, in order to enhance the accuracy of the judgment of the verification terminal, the verification terminal may also prompt the user to perform face authentication again, so as to ensure the accuracy of the authentication result, and the specific steps are as follows:
if not, the verification terminal sends out a verification failure prompt to remind the user to perform identity verification again.
In addition, because the movement of the two eyes is basically synchronous during movement, for this reason, the method for calculating the focal distance between the two eyes is adopted in the application to assist in forming the eyeball movement tracks generated by the rotation of the two eyes respectively, and the specific steps are as follows:
and S230, two cameras used for respectively shooting eyes of the two eyes of the user are arranged on the verification terminal, the verification terminal obtains the focal length between the two eyes by analyzing the positions of the focuses of the shot eyes, and the formation of the eyeball motion track is assisted according to the focal length between the two eyes.
Specifically, a unique normal is specified through the focal length between the two eyes, the normal is perpendicular to a connecting line between the focal points of the two eyes, the eyeball movement tracks formed by the respective motions of the two eyes are symmetrical to the normal, and the eyeball movement tracks formed by the rotation of the two eyes can be obtained simultaneously according to the characteristics of a symmetrical graph.
In the step S230: the method for assisting the formation of the eyeball motion trail according to the focal length between the two eyes specifically comprises the following steps:
s231, the movement of the eyeballs of the two eyes of the user corresponds to an eyeball movement track respectively, and the eyeball movement tracks of the two eyes are symmetrical by taking the focal length of the two eyes as a boundary. Specifically, when the eye movement trajectories of both eyes are symmetrical, it is described that the eyes move synchronously, and any one of the movement trajectories can be used for comparison with the verification line.
And S232, when the eyeball motion tracks of the two eyes are asymmetric, the verification terminal combines the focal length and the focal point of the two eyes and the area of the iris and the sclera in the shot eyeball picture, and adjusts the other eyeball motion track by taking one eyeball motion track as a reference according to a similar principle.
The specific implementation mode of adjusting one eyeball motion track as a reference comprises the following steps:
the focal position of machine vision positioning and the area ratio of the sclera and the iris in the real eyeball are calculated through a geometric algorithm, one eyeball motion track in two eyeball motion tracks is analyzed to accord with the focal point and the change rule of the areas of the sclera and the iris, the selected eyeball motion track is taken as a reference, and the position of the focal point on the other eyeball motion track is adjusted according to a symmetry axis formed by all focal lengths, so that the two eyeball motion tracks are symmetrical.
After analyzing the focal position, the focal length of the eyes, and the area change of the iris and the sclera, obtaining a real eyeball motion track, comparing the eyeball motion track with a verification line in a verification terminal, further verifying the identity of a user, and specifically executing the following steps when verifying and judging:
and S250, the verification terminal compares the eyeball motion track with the verification line and judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value.
Specifically, the eyeball movement track is formed by the user rotating the eyeball according to a preset rule, and the eyeball of the user can be shown to be real by comparing the eyeball movement track formed by the user rotating the eyeball with a verification line used for verifying the authenticity of the eyeball of the user in the verification terminal, and the eyeball is rotated on the site provided with the verification terminal for identity verification.
S300, when the similarity of the eyeball motion track and a verification line corresponding to the verification line reaches a similarity threshold value, and meanwhile, a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed.
The verification line may be a combination of 26 letters or a combination of arabic numerals from 1 to 10, which is described below only by way of example, and it is understood that the combination of 26 letters is similar to the principle of arabic numeral combination.
The mode that the verification terminal generates the verification line according to the preset rule specifically comprises the following steps:
the verification terminal calculates the product of the current date and the month according to the current date and the month;
when the verification terminal receives an identity verification instruction of a to-be-verified person, the verification terminal acquires the current hour time, and the hour time is increased on the basis of the product of the current date and the month to form a verification line and store the verification line in the verification terminal.
Here, specific values are exemplified, for example, when the current time is 12 months and 10 days in 2019 and 10:00, the way of calculating the verification line is as follows: 12 x 10-120, 120+ 10-130.
In the preferred embodiment of the present invention, the ones and tens of the above numerical values are taken as two verification lines, and when the two digits are less than two, the ones are complemented by "0" in the tens, it should be understood that the complemented numbers in the tens are not limited, and any numbers can be complemented in the tens by the user definition, and any numbers can be complemented in the tens randomly by the system.
Meanwhile, in order to enhance the accurate positioning of the current hour time, when the verification line is formed by the verification terminal, a tolerance algorithm is added, and the obtained verification line can be adopted by the verification terminal by default when the time close to the moment is reached. Here, specific times are exemplified: when the user makes eyeball recording during 20:58:00 to 21:02:00, the data of the hour hand is approved by the system no matter whether 20 or 21 is selected, and the verification line obtained by taking 20 or 21 as the addend of the product of the date and the month can be adopted by the verification terminal.
The method for calculating the similarity threshold comprises the following steps:
and the verification terminal performs numerical analysis on the eyeball motion track, and calculates the numerical difference between the verification line and the eyeball motion track to obtain the similarity threshold. Specifically, the verification line is a two-digit number, when the eyeball movement locus is analyzed, the arabic number and the eyeball movement locus are compared to find out the same number, when the found number is the same as the number of the verification line, namely, the difference of the numbers is 0, the eyeball verification is successful, the similarity threshold is 0, and similarly, the eyeball verification is unsuccessful as long as the similarity threshold is not 0.
When the eyeball motion trail is input into the verification terminal, the following steps are executed (understandably, a plurality of verification lines correspond to a plurality of eyeball motion trails, and when the verification lines are two digits, the motion trail corresponds to two tracks):
the verification terminal receives an input starting instruction sent by a user, captures a focus of eyeballs of the user according to the input starting instruction, carries out focusing processing, and prompts the user to send a first input instruction after focusing is completed.
And the verification terminal shoots a first eyeball movement track generated by the first rotation of the eyeballs of the user according to the first input instruction and prompts the user to send a second input instruction.
And the verification terminal shoots a second eyeball movement track generated by the second rotation of the eyeballs of the user according to the second input instruction and prompts the user to send a third input instruction.
And circulating the steps until the verification terminal shoots the last eyeball motion track generated by the last eyeball rotation of the user.
And the verification terminal locks the end point of the last eyeball movement track to finish the input of all the eyeball movement tracks of the user.
Specifically, for example, as shown in fig. 3, fig. 3 is a flowchart of inputting an eye movement trajectory according to this embodiment.
S10, blinking once by the user, and starting the verification terminal;
s20, the verification terminal uses AI auxiliary focusing to focus the captured focus, and after focusing is successful, a prompt tone or screen flicker or screen character prompt and the like are generated;
s30, the user blinks once, and the verification terminal starts to input a first eyeball motion track;
s40, the user blinks twice quickly, and the verification terminal starts to input a second eyeball motion track;
s50, after the recording is finished, the track point is locked at the last position and kept 1S;
and S60, finishing the recording.
In a preferred embodiment of the present invention, when performing the authenticity identification of the eyeball of the user, a step stealth algorithm of the Z axis may be further added, and it can be understood that the eyeball of the user is a three-dimensional sphere, and when the eyeball rotates, the eyeball contracts with respect to the orbit, and at this time, the contraction of the eyeball according to the center of the viewing distance is represented by the Z axis direction.
Specifically, when the authentication terminal captures a face image and an eye movement locus of the user, the method further includes:
verifying the visual distance of the terminal when positioning the eyeball of the user, and judging whether the change of the positioned visual distance is matched with the movement track of the eyeball of the user;
if the eyeball is matched with the verification line, the eyeball of the user is real, and the verification terminal judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not;
if not, the user authentication fails, and the authentication is quitted.
As shown in fig. 4, when the user identity is verified by using the Z-axis advanced stealth algorithm, the verification terminal performs the following steps:
and Z100, the verification terminal receives an identity verification instruction sent by the user and generates a verification line for verifying the authenticity of the eyeballs of the user according to the identity verification instruction and a preset rule.
And Z200, shooting the facial image and the eyeball motion track of the user by the verification terminal, and positioning the visual distance of the user when the eyeball rotates.
Z300, the verification terminal judges whether the change of the positioned sight distance is matched with the eyeball motion track of the user or not;
z310, if the verification is not matched, sending a verification failure prompt, initializing the verification step, and repeating the step Z100;
z400, if the eyeball motion trail and the verification line are matched, judging whether the similarity of the eyeball motion trail and the verification line reaches a similarity threshold value;
z410, if not, sending out a verification failure prompt, initializing a verification step, and repeating the step Z100;
z500, if yes, traversing the cloud database to judge whether a face verification image matched with the shot face image exists;
z510, if not, sending out a verification failure prompt, initializing a verification step, and repeating the step Z100;
and Z600, if so, the user passes the authentication.
It should be understood that the above steps Z300, Z400 and Z500 are not sequential, and can be performed simultaneously or in steps, and the above is only a specific embodiment of the present solution, and is used to illustrate the specific application of the Z-axis step hiding algorithm in the authentication terminal when applied to the identity authentication.
The specific calculation and verification method is similar to the above-mentioned area ratio of the focal point, iris and sclera, and is not described herein again.
The invention also discloses a verification terminal, wherein, as shown in fig. 5, the verification terminal comprises a processor 10 and a memory 20 connected with the processor 10, the memory 20 stores a configuration program of the high-security face recognition method, and the configuration program of the high-security face recognition method is executed by the processor 10 to realize the high-security face recognition method; as described above.
The invention also discloses a storage medium, wherein the storage medium stores a computer program which can be executed for realizing the high-security face recognition method; as described above.
In summary, the present invention discloses a high-security face recognition method, a verification terminal and a storage medium, wherein the method includes: the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule; the verification terminal shoots a face image and an eyeball motion track of a user and judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not; traversing a cloud database to judge whether a face verification image matched with the shot face image exists or not; and when the similarity of the eyeball motion track and the verification line corresponding to the verification line reaches a similarity threshold value, and meanwhile, a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed. The verification terminal generates a verification line according to a preset rule, the user rotates an eyeball according to the preset rule, the verification terminal shoots an eyeball movement track of the user and simultaneously acquires the face characteristics of the user, the user identity verification is passed under the condition that the face characteristics and the eyeball movement track are both recognized, the safety level of the verification terminal for face recognition is improved, the real face can be judged by recognizing the face through double-layer verification, and the high safety and accuracy of the face recognition are ensured; meanwhile, when the authenticity of the eyeball is judged, the judgment is carried out according to the eyeball motion track, the area ratio of the iris and the sclera when the eyeball of the user really rotates is calculated in advance through a geometric algorithm, then the motion track analyzed by the verification terminal is judged, measures are reasonably taken to carry out the fitting of a new eyeball motion track when the iris and the sclera are not related, the focal length between two eyes is utilized to carry out auxiliary fitting when the eyeball motion track is fitted, the authenticity and the validity of the fitted eyeball motion track are favorably ensured, and the verification terminal is convenient to carry out eyeball authentication.
It is to be understood that the invention is not limited to the examples described above, but that modifications and variations may be effected thereto by those of ordinary skill in the art in light of the foregoing description, and that all such modifications and variations are intended to be within the scope of the invention as defined by the appended claims.
Claims (11)
1. A high-density face recognition method is characterized by comprising the following steps:
the verification terminal receives an identity verification instruction sent by a user and generates a verification line for verifying the authenticity of eyeballs of the user according to the identity verification instruction and a preset rule;
the verification terminal shoots a face image and an eyeball motion track of a user, judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not, and traverses a cloud database to judge whether a face verification image matched with the shot face image exists or not;
and when the similarity between the eyeball motion track and the verification line reaches a similarity threshold value and a face verification image matched with the shot face image exists in the cloud database, the identity verification of the user is passed.
2. The high-security face recognition method according to claim 1, wherein the verification terminal captures an eye movement trajectory of the user, and determines whether the similarity between the eye movement trajectory and the verification line reaches a similarity threshold, specifically comprising:
the verification terminal shoots pictures of eyeballs of the user according to a time axis and connects all eyeball focuses in the shot pictures to form an eyeball motion track;
and the verification terminal compares the eyeball motion track with the verification line and judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not.
3. The high-security face recognition method according to claim 2, wherein the verification terminal takes pictures of the user's eyes according to a time axis and connects all the eye focuses in the taken pictures to form an eye movement track, and further comprising:
the verification terminal judges whether the areas of the irises and the sclera in the eyeballs corresponding to all the focuses on the eyeball motion track formed by connecting the verification terminal with the focuses are close to the areas of the irises and the sclera in the shot picture or not according to the shot picture;
if the two points are close to each other, the verification terminal compares the eyeball motion track formed by connecting all the focuses with the verification line;
if the positions of the iris and the sclera in the shot picture are not similar, the verification terminal fits a new eyeball motion track according to the areas of the iris and the sclera in the shot picture and the position of the focus on the eyeball motion track, and the new eyeball motion track is compared with the verification line.
4. The high-density face recognition method according to claim 3, wherein the verification terminal is provided with two cameras for respectively shooting eyes of two eyes of the user, the verification terminal obtains a focal length between the two eyes by analyzing the positions of the focuses of the shot eyes, and the formation of the eye movement track is assisted according to the focal length between the two eyes.
5. The high-density face recognition method according to claim 4, wherein the verification terminal assists in forming the eye movement trajectory according to the focal length between the two eyes, and specifically comprises:
the movement of the eyeballs of the two eyes of the user corresponds to an eyeball movement track respectively, and the eyeball movement tracks of the two eyes are symmetrical by taking the focal length of the two eyes as a boundary;
when the eyeball motion tracks of the two eyes are asymmetric, the verification terminal combines the focal length and the focal point of the two eyes and the area of the iris and the sclera in the shot eyeball picture, and adjusts the other eyeball motion track by taking one eyeball motion track as a reference according to a similar principle.
6. The high-security face recognition method according to claim 1, wherein the way for the verification terminal to generate the verification line according to the preset rule is specifically:
the verification terminal calculates the product of the current date and the month according to the current date and the month;
when the verification terminal receives an identity verification instruction of a to-be-verified person, the verification terminal acquires the current hour time, and the hour time is increased on the basis of the product of the current date and the month to form a verification line and store the verification line in the verification terminal.
7. The high-security face recognition method according to claim 6, wherein the similarity threshold is calculated by:
and the verification terminal performs numerical analysis on the eyeball motion track, and calculates the numerical difference between the verification line and the eyeball motion track to obtain the similarity threshold.
8. The high-security face recognition method according to claim 1, wherein a verification terminal captures a face image and an eye movement track of a user, and determines whether the similarity between the eye movement track and the verification line reaches a similarity threshold, wherein the mode of entering the eye movement track of the user by the verification terminal comprises:
the verification terminal receives an input starting instruction sent by a user, captures a focus of eyeballs of the user according to the input starting instruction, carries out focusing processing, and prompts the user to send a first input instruction after focusing is finished;
the verification terminal shoots a first eyeball movement track generated by the first rotation of the eyeballs of the user according to the first input instruction, and prompts the user to send a second input instruction;
the verification terminal shoots a second eyeball movement track generated by the second rotation of the eyeballs of the user according to the second input instruction, and prompts the user to send a third input instruction;
the steps are repeated until the verification terminal shoots the last eyeball movement track generated by the last rotation of the eyeballs of the user;
and the verification terminal locks the end point of the last eyeball movement track to finish the input of all the eyeball movement tracks of the user.
9. The high-security face recognition method according to claim 1, wherein the authentication terminal captures a facial image and an eye movement trajectory of the user, and further comprises:
verifying the visual distance of the terminal when positioning the eyeball of the user, and judging whether the change of the positioned visual distance is matched with the movement track of the eyeball of the user;
if the eyeball is matched with the verification line, the eyeball of the user is real, and the verification terminal judges whether the similarity between the eyeball motion track and the verification line reaches a similarity threshold value or not;
if not, the user authentication fails, and the authentication is quitted.
10. An authentication terminal comprising a processor and a memory connected to the processor, wherein the memory stores a configuration program of a high-security face recognition method, and the configuration program of the high-security face recognition method realizes the high-security face recognition method according to any one of claims 1 to 9 when executed by the processor.
11. A storage medium, characterized in that the storage medium stores a computer program executable for implementing the high-security face recognition method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010285942.4A CN111475791B (en) | 2020-04-13 | 2020-04-13 | High-security face recognition method, verification terminal and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010285942.4A CN111475791B (en) | 2020-04-13 | 2020-04-13 | High-security face recognition method, verification terminal and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111475791A true CN111475791A (en) | 2020-07-31 |
| CN111475791B CN111475791B (en) | 2023-01-24 |
Family
ID=71752253
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010285942.4A Active CN111475791B (en) | 2020-04-13 | 2020-04-13 | High-security face recognition method, verification terminal and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111475791B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120207357A1 (en) * | 2010-08-06 | 2012-08-16 | Honeywell International Inc. | Ocular and iris processing system and method |
| US20160019420A1 (en) * | 2014-07-15 | 2016-01-21 | Qualcomm Incorporated | Multispectral eye analysis for identity authentication |
| CN105279492A (en) * | 2015-10-22 | 2016-01-27 | 北京天诚盛业科技有限公司 | Iris identification method and device |
| US20160112414A1 (en) * | 2014-10-15 | 2016-04-21 | Utechzone Co., Ltd. | Network authentication method and system based on eye tracking procedure |
| US20170193213A1 (en) * | 2016-01-04 | 2017-07-06 | Utechzone Co., Ltd. | Eye movement traces authentication system, method, and non-transitory computer readable medium, the same which integrate with face recognition and hand recognition |
| CN107016270A (en) * | 2015-12-01 | 2017-08-04 | 由田新技股份有限公司 | Dynamic graphic eye movement authentication system and method combining face authentication or hand authentication |
-
2020
- 2020-04-13 CN CN202010285942.4A patent/CN111475791B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120207357A1 (en) * | 2010-08-06 | 2012-08-16 | Honeywell International Inc. | Ocular and iris processing system and method |
| US20160019420A1 (en) * | 2014-07-15 | 2016-01-21 | Qualcomm Incorporated | Multispectral eye analysis for identity authentication |
| US20160112414A1 (en) * | 2014-10-15 | 2016-04-21 | Utechzone Co., Ltd. | Network authentication method and system based on eye tracking procedure |
| CN105279492A (en) * | 2015-10-22 | 2016-01-27 | 北京天诚盛业科技有限公司 | Iris identification method and device |
| CN107016270A (en) * | 2015-12-01 | 2017-08-04 | 由田新技股份有限公司 | Dynamic graphic eye movement authentication system and method combining face authentication or hand authentication |
| US20170193213A1 (en) * | 2016-01-04 | 2017-07-06 | Utechzone Co., Ltd. | Eye movement traces authentication system, method, and non-transitory computer readable medium, the same which integrate with face recognition and hand recognition |
Non-Patent Citations (1)
| Title |
|---|
| 许菲菲: "基于人眼识别的可穿戴设备身份认证", 《网络空间安全》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111475791B (en) | 2023-01-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6775397B1 (en) | Method and apparatus for user recognition using CCD cameras | |
| US9619723B1 (en) | Method and system of identification and authentication using facial expression | |
| US8280120B2 (en) | Fraud resistant biometric financial transaction system and method | |
| US8184914B2 (en) | Method and system of person identification by facial image | |
| KR101242390B1 (en) | Method, apparatus and computer-readable recording medium for identifying user | |
| Killioğlu et al. | Anti-spoofing in face recognition with liveness detection using pupil tracking | |
| US8750580B2 (en) | Comparing device | |
| JP4760049B2 (en) | Face authentication device, face authentication method, electronic device incorporating the face authentication device, and recording medium recording the face authentication program | |
| CN1160446A (en) | Automated, non-invasive iris reconition system and method | |
| CN111402480A (en) | Visitor information management method, device, system, equipment and storage medium | |
| CN112507889A (en) | Method and system for verifying certificate and certificate holder | |
| CA2833740A1 (en) | Method of generating a normalized digital image of an iris of an eye | |
| KR101954763B1 (en) | Face recognition access control apparatus and operation method thereof | |
| KR101821144B1 (en) | Access Control System using Depth Information based Face Recognition | |
| US11663853B2 (en) | Iris authentication device, iris authentication method, and recording medium | |
| CN113614731A (en) | Authentication verification using soft biometrics | |
| CN111666835A (en) | Face living body detection method and device | |
| KR102146552B1 (en) | Non face to face authentication system | |
| WO2020065954A1 (en) | Authentication device, authentication method, and storage medium | |
| JP2006085289A (en) | Facial authentication system and facial authentication method | |
| CN111475791B (en) | High-security face recognition method, verification terminal and storage medium | |
| CN112257507A (en) | Method and device for judging distance and human face validity based on human face interpupillary distance | |
| US20220067135A1 (en) | Method and System for Authenticating an Occupant Within an Interior of a Vehicle | |
| CN115019364A (en) | Identity authentication method and device based on face recognition, electronic equipment and medium | |
| CN114299569A (en) | Safe face authentication method based on eyeball motion |
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 |