CN110942042B - Three-dimensional handwritten signature authentication method, system, storage medium and equipment - Google Patents
Three-dimensional handwritten signature authentication method, system, storage medium and equipment Download PDFInfo
- Publication number
- CN110942042B CN110942042B CN201911213154.8A CN201911213154A CN110942042B CN 110942042 B CN110942042 B CN 110942042B CN 201911213154 A CN201911213154 A CN 201911213154A CN 110942042 B CN110942042 B CN 110942042B
- Authority
- CN
- China
- Prior art keywords
- curve
- dimensional
- transformation
- signature
- template
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000009466 transformation Effects 0.000 claims description 108
- 230000033001 locomotion Effects 0.000 claims description 35
- 238000013519 translation Methods 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 13
- 230000002452 interceptive effect Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000010845 search algorithm Methods 0.000 claims description 7
- 230000001131 transforming effect Effects 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000002068 genetic effect Effects 0.000 description 4
- 230000003238 somatosensory effect Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000006698 induction Effects 0.000 description 3
- 241001422033 Thestylus Species 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 206010027175 memory impairment Diseases 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001172 regenerating effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/30—Writer recognition; Reading and verifying signatures
- G06V40/37—Writer recognition; Reading and verifying signatures based only on signature signals such as velocity or pressure, e.g. dynamic signature recognition
- G06V40/394—Matching; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention relates to a three-dimensional handwritten signature authentication method, a system, a storage medium and equipment, which realize signature authentication by acquiring and identifying a three-dimensional signature track, convert the three-dimensional signature track into a two-dimensional curve to be tested on a maximum projection plane, and judge the similarity between the curve to be tested and a template curve by calculating the similar distance after the curve to be tested is matched with the template curve, thereby confirming whether the three-dimensional signature track is a real signature of a user corresponding to the template curve. Compared with the prior art, the invention has the characteristics of high safety and difficult counterfeiting.
Description
Technical Field
The invention relates to the field of identity recognition, in particular to a three-dimensional handwritten signature authentication method, a system, a storage medium and equipment.
Background
With the development of information technology, biometric identity recognition methods including handwritten signatures are widely used, and handwritten signatures become a main method of identity recognition due to their high uniqueness and reliability. The characteristic information of the hand-written signature is compared with the real signature sample by a computer to identify the authenticity of the hand-written signature, and the principle is that the signature of each person is unique and can not be changed randomly within a period of time. The handwritten signature has the characteristics of non-forgetfulness, naturalness, shareability and certain relative instability. The handwritten signature recognition system has to have enough robustness and give consideration to the system performance, and the mature signature recognition system has wide application prospects, such as the fields of finance, securities and intelligent commerce, enterprise resource management systems, office automation systems, intelligent government affairs and the like.
At present, equipment such as a handwriting pad, a mobile phone and a tablet computer can be used for conveniently collecting handwriting, and identity authentication is carried out through the handwriting. However, the signatures collected by devices such as a handwriting pad, a mobile phone, a tablet personal computer and the like are two-dimensional signature handwriting, and the two-dimensional signature handwriting (mainly a signature image) is easy to collect and imitate, so that the security is low, and the two-dimensional signature handwriting is easy to forge, thereby causing the loss of the user in terms of economy and reputation.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a three-dimensional handwritten signature authentication method which is not easy to forge and has high safety.
A three-dimensional handwritten signature authentication method comprises the following steps:
acquiring a three-dimensional signature track corresponding to a three-dimensional Cartesian coordinate system data format;
respectively carrying out rotation transformation on the three-dimensional signature track along an X axis and a Y axis of an XYZ reference system, selecting a target X axis rotation angle and a target Y axis rotation angle when the projections of the three-dimensional signature track on an XZ plane and a YZ plane are minimum in the rotation transformation process of the three-dimensional signature track, and obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured when the three-dimensional signature track is rotated to the target X axis rotation angle and the target Y axis rotation angle;
acquiring a template curve corresponding to the target user information;
segmenting the template curve and the curve to be compared, wherein each segment of the template curve has a corresponding interval to be compared on the curve to be compared; carrying out similarity transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and calculating the average distance between the curve to be tested in the interval to be compared and the transformation curve segment as the similarity distance;
and carrying out weighted average on the similar distances corresponding to the one or more than one interval to be compared to obtain an average similar distance, and if the average similar distance is smaller than a given threshold value, judging that the three-dimensional signature track is a real signature matched with the target user information.
In one embodiment, the step of performing rotation transformation on the three-dimensional signature track along the X axis and the Y axis of an XYZ reference system respectively, and in the rotation transformation process of the three-dimensional signature track, selecting a target X-axis rotation angle and a target Y-axis rotation angle when the projections of the three-dimensional signature track on the XZ plane and the YZ plane are minimum, and acquiring a projection curve of the three-dimensional signature track on the XY plane as a curve to be measured when the three-dimensional signature track is rotated to the target X-axis rotation angle and the target Y-axis rotation angle includes:
step S201: carrying out X-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein the rotation transformation formula is as follows:
theta is the rotation angle of the X-axis rotation transformation and belongs to the angle range interval of [ -90, 90 ];
step S202: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each X-axis rotation transformation, and selecting the angle theta corresponding to the one-time rotation transformation with the minimum variance as a target X-axis rotation angle;
step S203: carrying out Y-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein the rotation transformation formula is as follows:
step S204: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each Y-axis rotation transformation, and selecting the angle corresponding to the rotation transformation with the minimum variance
As a target Y-axis rotation angle;
step S205: taking the value of theta as the target X-axis rotation angle,
when the value is taken as the rotation angle of the target Y axis, the transformation matrix T = T 1 T 2 And obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured.
In one embodiment, the step of obtaining the template curve corresponding to the target user information includes:
acquiring a template three-dimensional signature track corresponding to a three-dimensional Cartesian coordinate coefficient data format;
and executing the steps S201-S205 on the template three-dimensional signature track to obtain a projection curve of the template three-dimensional signature track on an XY plane as a template curve.
In one embodiment, the step of performing similar transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and calculating the average distance between the curve to be tested in the interval to be compared and the transformation curve segment as the similar distance includes:
for a section [ C ] to be compared with an initial position offset of t 1 ,C 2 ]Carrying out similarity transformation on the template curve according to the transformation matrix T;
the transformation matrix T is
a. d is scaling in X and Y directions respectively; l and m are translation distances in X and Y directions respectively;
searching subsection interval [ R ] on template curve by using intelligent search algorithm 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]And the corresponding transformation curve segment with the maximum coincidence degree of the curve segment to be detected. The reference comparison interval [ R 1 ,R 2 ]Corresponding transformation curve segment F A Comprises the following steps:
F A ＝{(x 1 ,y 1 ),…(x 2 ,y 2 ),…(x M ,y M )}
the interval [ C ] to be compared 1 ,C 2 ]Corresponding curve segment F to be measured B Comprises the following steps:
F B ＝{(x′ 1 ,y′ 1 ),…(x′ 2 ,y′ 2 ),…(x′ N ,y′ N )}
wherein M and N represent a transformation curve segment F A And the curve segment F to be measured B The number of feature points of;
sampling a plurality of sampling points at equal intervals from the transformation curve segment and the curve segment to be detected, and calculating the average distance between the sampling points at the corresponding positions as the similar distance between the transformation curve segment and the curve segment to be detected corresponding to the interval t to be compared:
wherein | | | is a distance norm, [ |]Represents rounding, Q is the number of sample points, F Ai F Bi Respectively a transformation curve segment and a sampling point of a curve segment to be detected.
In one embodiment, the searching by using the intelligent search algorithm obtains the template curve segment interval [ R 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]The step of transforming the curve segment with the maximum coincidence degree of the corresponding curve segment to be tested comprises the following steps:
s421: randomly generating an initial population P (j); wherein j =0;
s422: randomly generating a new individual in the neighborhood of the individual according to the values of the scaling a in the X direction, the scaling d in the Y direction, the translation distance l in the X direction, the translation distance m in the Y direction and the initial position offset t of the interval to be compared in the individual;
s423: calculating the fitness values of all parents and genetically-generated offspring, selecting S offspring according to the fitness values to form a new population P (j + 1) and replacing the population P (j);
s424: when the maximum iteration times are reached and the loop iteration reaches the maximum generation number, stopping, and selecting the transformation curve segment with the maximum coincidence degree with the curve segment to be detected for output; otherwise, go to step S422.
In one embodiment, the step of obtaining the template curve corresponding to the target user information includes:
and receiving and authenticating the input target user information, and acquiring a template curve corresponding to the target user information in the server after the authentication is passed.
The invention aims to overcome the defects of the prior art and provide a three-dimensional handwritten signature authentication system which is not easy to forge and has high safety.
A three-dimensional handwritten signature authentication system comprising: the system comprises a body sensing device, a database server and an authentication server; the motion sensing device is used for collecting a three-dimensional signature track, the database server is used for storing the template curve, and the authentication server is used for executing the three-dimensional handwritten signature authentication method;
the motion sensing device is used for capturing user gestures through a camera to acquire a three-dimensional signature track; or,
the motion sensing device is integrated on the user terminal and used for acquiring a three-dimensional signature track by detecting a spatial motion track of the user terminal.
In one embodiment, the three-dimensional handwritten signature authentication system further includes: and the interactive display equipment is used for receiving the input target user information and displaying the three-dimensional handwritten signature authentication result.
The present invention is directed to overcome the disadvantages and drawbacks of the prior art and to provide a computer-readable storage medium storing a computer program for executing the above-mentioned three-dimensional handwritten signature authentication method that is not easily forged and has high security.
A computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the three-dimensional handwritten signature authentication method of any of the preceding claims.
The invention aims to overcome the defects of the prior art and provide a computer device for storing and executing the three-dimensional handwritten signature authentication method which is difficult to forge and has high safety.
A computer device comprising a memory, a processor and a computer program stored in the memory and executable by the processor, the processor implementing the three-dimensional handwritten signature authentication method according to any of the preceding claims when executing the computer program.
After the three-dimensional handwritten signature authentication method, the three-dimensional handwritten signature authentication system, the storage medium and the equipment are adopted, the three-dimensional signature track of a user during signature can be detected through the somatosensory equipment, the three-dimensional signature track is different from two-dimensional handwriting in the traditional technology, the three-dimensional signature track is writing posture information determined by writing habits of the user during writing, a two-dimensional curve to be tested is obtained after the three-dimensional signature track expressing the writing posture information is rotated and transformed, then similarity comparison is carried out on the curve to be tested and a template curve corresponding to a target user, and the three-dimensional signature track is judged to be a real signature of the target user when the similarity is higher, so that the authentication result is more accurate, and the signature authentication safety is improved.
For a better understanding and practice, the invention is described in detail below with reference to the accompanying drawings.
Drawings
FIG. 1 is a flow chart of a method for authenticating a three-dimensional handwritten signature according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a trace of a user writing a simple character "A" in an embodiment of the present invention;
FIG. 3 is a flowchart of step S2 of a three-dimensional handwritten signature authentication method according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a process of transforming a template curve segment and matching the template curve segment with a curve segment to be tested according to an embodiment of the present invention;
FIG. 5 is a flowchart of step S4 of a three-dimensional handwritten signature authentication method in an embodiment of the present invention;
FIG. 6 is a flowchart of curve matching using a genetic algorithm in step S4 according to an embodiment of the present invention;
FIG. 7 is a schematic structural diagram of a three-dimensional handwritten signature authentication system in an embodiment of the present invention;
FIG. 8 is a schematic diagram of a three-dimensional handwritten signature authentication system according to another embodiment of the present invention;
fig. 9 is a schematic structural diagram of acquiring a three-dimensional handwritten signature track performed by a user through a gesture in space by using a camera in an active somatosensory method according to an embodiment of the present invention;
fig. 10 is a schematic diagram of a smart pen with a built-in sensor for acquiring a three-dimensional handwritten signature track by a passive body sensing method according to an embodiment of the present invention.
Detailed Description
Examples
Referring to fig. 1, the present invention provides a method for authenticating a three-dimensional handwritten signature, comprising the following steps:
step S1: and acquiring a three-dimensional signature track corresponding to a three-dimensional Cartesian coordinate system data format.
In this embodiment, a three-dimensional signature trajectory is acquired by an acquisition device and converted into a data format of a three-dimensional cartesian coordinate system, that is, a sequence of sample points represented by a three-dimensional vector in an XYZ coordinate system constitutes trajectory data.
The acquisition device may be either active or passive. The active collection means that a three-dimensional gesture track when a user signs is actively captured through collection equipment, for example, a gesture motion when the user signs can be shot through a camera, so that the three-dimensional signature track when the user signs is collected.
The passive acquisition refers to that the acquisition equipment is held by a user or moves along with the movement of a gesture when the user signs, and a three-dimensional signature track is passively acquired according to the movement track of the acquisition equipment.
For example, in an application scenario of a smart pen, a capture device may be integrated in the smart pen, and the capture device captures a motion trajectory of a fingertip or a pen tip of a user, that is, a "pen" captured by the capture device, to obtain a signature trajectory in a virtual space.
The acquisition equipment can adopt inertial sensing, optical sensing and inertial and optical combined sensing equipment, wherein an inertial sensing user needs to bind an inertial sensor or hold an inertial sensing pen by hand, and somatosensory data is obtained by utilizing real-time change of the sensor; the optical sensing device can be a fixed body sensing device or a photoelectric digital pen held by a user, the signature track is a space continuous curve, and the user can select the sensing mode of the body sensing device according to actual requirements. When the pen enters the induction space, gesture data acquisition is started, the motion track of the pen is recorded by the equipment, and when the pen leaves the induction space, gesture data acquisition is finished. In one embodiment, fixed gesture operation can also be used to confirm the start and stop of the signature data, such as "pen" pause when writing is started, "pen" pause again when writing is finished, or signature start when the hand is lifted, signature end when the hand is put down, and the like.
Preferably, before the three-dimensional signature track is identified, data points in the three-dimensional signature track can be screened, the motion speed of each data point in the signature track is calculated, the average speed is counted, and data points with the motion speed smaller than the average speed are removed; rejecting abnormal position points such as data points with the distance between two data points exceeding a set range or the quantity of the data points being lower than a set number and lacking continuity in the signature data; if the distance between the two data points exceeds a set range, the signature position is possibly mutated; the data points with the data point quantity lower than the set number and lacking continuity can be a small number of data points captured by accidentally reentering the sensing space after the user signature is written and when the hand leaves the sensing space; the presence of this data point affects the accuracy of subsequent signature authentications.
And carrying out difference on the sampled ordered point set by using equal time intervals to realize resampling. The difference method can adopt various methods such as linearity, spline and the like.
As shown in fig. 2, it is a schematic diagram of a trajectory of a user executing a "pen" to write a character "a", where a trajectory BC is a signature trajectory, a trajectory AB is a motion trajectory of the stylus pen entering the sensing space, and a trajectory CD is a motion trajectory of the stylus pen leaving the sensing space. Because the user stops obviously at the point B and the point C, the key positions and the time of the point B and the point C can be obtained by counting the minimum track motion speed point, and the track BC is saved. In other embodiments, instead of processing the B and C key points, the trace AD may be saved as the entire signature trace, and then the influence of the AB and CD parts may be eliminated by comparing the trace AD with the template curve.
In one embodiment, before the three-dimensional signature track and the template signature track are subjected to rotation transformation along an X axis and a Y axis respectively, preprocessing including abnormal point elimination, smoothing and filtering is performed on the three-dimensional signature track and the template signature track, the abnormal points refer to obvious pause or special states in the signature track, and effective signature data is segmented and extracted on the three-dimensional signature track and the template signature track, so that the accuracy of subsequent signature authentication is improved, the comparison data amount is reduced, and the efficiency is improved.
Step S2: and respectively carrying out rotation transformation on the three-dimensional signature track along an X axis and a Y axis of an XYZ reference system, selecting a target X axis rotation angle and a target Y axis rotation angle when the projections of the three-dimensional signature track on an XZ plane and a YZ plane are minimum in the rotation transformation process of the three-dimensional signature track, and obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured when the three-dimensional signature track is rotated to the target X axis rotation angle and the target Y axis rotation angle.
As shown in fig. 3, the steps specifically include:
step S201: carrying out X-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein the rotation transformation formula is as follows:
theta is the rotation angle of the X-axis rotation transformation and belongs to the angle range interval of [ -90, 90 ];
step S202: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each X-axis rotation transformation, and selecting the angle theta corresponding to the one-time rotation transformation with the minimum variance as a target X-axis rotation angle;
step S203: carrying out Y-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein a rotation transformation formula is as follows:
step S204: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each Y-axis rotation transformation, and selecting the angle corresponding to the rotation transformation with the minimum variance
As a target Y-axis rotation angle;
step S205: taking the value of theta as the target X-axis rotation angle,
when the value is the target Y-axis rotation angle, the transformation matrix T = T 1 T 2 And obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured.
Using transformation matrix T = T 1 T 2 Converting the three-dimensional signature track and the template signature track into a two-dimensional curve on an XY plane, and after obtaining a curve to be tested, carrying out coordinate translation and rotation transformation on the signature track data to ensure that the signature track is on a plane approximately perpendicular to a Z axis, the track center is a coordinate origin, and the X coordinate and the Y coordinate correspond to a writing coordinate of the signature data on the two-dimensional plane; transformed matrix T = T 1 T 2 Then, the signature writing can be approximately considered to be written on a plane parallel to the XY plane, and the similarity of the three-dimensional curve can be converted into a two-dimensional curve. Preferably, the coordinate values of the curve to be measured can be normalized to be [ -1,1]In the process of matching the subsequent curves, the search space in the subsequent curve matching process is reduced, and the search speed is improved.
And step S3: and acquiring a template curve corresponding to the target user information.
The user identity information can be authenticated in advance, and the user template signature track can be obtained from the server according to the authentication result. The server is provided with a user signature database, the user signature database comprises at least one user ID and at least one corresponding signature track, after the user identity information is authenticated, one signature track is extracted from the user signature database according to the user ID, and the extraction process can be random extraction or appointed extraction.
For example, in an application scenario of electronic payment, a user inputs an account password first to complete identity authentication of the account password, but the authentication can only confirm that the account password is matched with the account password, but the account password of the user is leaked to other people and the user is stolen to swipe a card by other people, so that the user providing the account password and the account password cannot be confirmed to be a real user corresponding to the account. At the moment, an account with successful password authentication can be obtained, a prestored template curve can be obtained in the database, the template curve can be input by a real user of the account during account opening registration, then the curve to be tested is compared with the template curve, if the similarity reaches a threshold value, the three-dimensional signature track corresponding to the curve to be tested is determined to be in accordance with the writing habit of the real user of the account, and therefore on the basis of account password authentication, the authenticity of an account provider is further authenticated, and therefore safety is improved.
It should be noted that the pre-stored template curve may be a two-dimensional curve mapped on the XY plane, or a pre-stored template three-dimensional signature track, and then the template three-dimensional signature track corresponding to the three-dimensional cartesian coordinate data format is obtained; and executing the steps S201-S205 on the template three-dimensional signature track to obtain a projection curve of the template three-dimensional signature track on an XY plane as a template curve. And for the template curve, the coordinate values can be normalized to be between [ -1,1] so as to reduce the search space in the subsequent curve matching process and improve the search speed.
And step S4: segmenting the template curve and the curve to be compared, wherein each segment of the template curve has a corresponding interval to be compared on the curve to be compared; and performing similar transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and calculating the average distance between the curve to be tested in the interval to be compared and the transformation curve segment as the similar distance.
In this embodiment, for a template curve with a length M, the template curve is uniformly divided into K segments by segment length Mk, and the segment interval is:
for a curve to be measured with the length of N, dividing the curve to be measured into K sections, wherein the interval to be compared of each section is as follows:
and in each interval to be compared, performing similar transformation on the template curve to obtain a transformation curve segment, enabling the coincidence degree of the transformation curve segment and the curve segment to be tested to be maximum, and calculating the average distance between the transformation curve segment and the curve segment to be tested to obtain the similar distance in the comparison interval.
The process of transforming the template curve and matching the curve segment to be measured is shown in FIG. 4, wherein [ R ] 1 ,R 2 ]Is a segmented interval of the template curve, [ C ] 1 ,C 2 ]Respectively, the intervals to be compared of the curves to be measured. And searching the optimal similarity transformation parameter in a similarity transformation space to minimize the similarity distance by calculating the minimum value of the differences between the two curves after the similarity transformation, and judging that the transformation of the template curve segment and the coincidence degree of the template curve segment to be detected are maximum, wherein H is the height of the template curve, L is the width of the template curve, a and d are scaling ratios in the X and Y directions respectively, and L and m are translation in the X and Y directions respectively.
As shown in fig. 5, the step of performing similar transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and calculating an average distance between the curve to be tested in the interval to be compared and the transformation curve segment as a similar distance, includes:
step S41: for a section [ C ] to be compared with a starting position offset of t 1 ,C 2 ]Performing similar transformation on the template curve by using a transformation matrix T according to the X-direction scaling a, the Y-direction scaling d, the X-direction translation distance l and the Y-direction translation distance m;
the transformation matrix T is
a. d is scaling in X and Y directions respectively; l and m are translation distances in X and Y directions respectively;
the transformation matrix T is obtained according to a two-dimensional plane basis transformation matrix, and in this embodiment, the transformation matrix T is a simple transformation matrix including only stretching and translation transformations without consideration of the Z axis and the rotational change:
wherein a and d are scaling ratios in X and Y directions respectively; b. c is the staggered cutting in the X direction and the Y direction respectively; a. b, c and d are combined into rotary transformation; l and m are respectively translation in X and Y directions; p and q are perspective transformation, generally do not consider the influence of the perspective transformation, and can be considered as 0; and s is a full scale transform coefficient.
Step S42: searching to obtain template curve segment interval R by using intelligent search algorithm such as genetic algorithm 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]Corresponding transformation curve section F with maximum coincidence degree of curve sections to be detected A The genetic algorithm takes the inverse of the similarity distance as the fitness function f. And simultaneously, the parameters a, d, l, m and t are required to meet certain constraint conditions, and the constraint conditions and the maximum generation number of the genetic algorithm are set according to practical application. Randomly generating S individuals P = { P = { P = } 1 ,P 2 ,…,P S Therein ofEach individual corresponds to the matching distance of a reference signature and a comparison signature segment, the reciprocal of the matching distance is the fitness of the individual, and the larger the fitness is, the smaller the distance between the reference signature participating in matching and the comparison signature segment is, namely, the better the individual is. The number of iterations may be set according to the initial individuals, for example, when 20 individuals are initially set, 100 iterations are set to converge left and right.
The transformation curve segment F A Comprises the following steps:
F A ＝{(x 1 ,y 1 ),…(x 2 ,y 2 ),…(x M ,y M )}
the curve segment F to be measured B Comprises the following steps:
F B ＝{(x′ 1 ,y′ 1 ),…(x′ 2 ,y′ 2 ),…(x′ N ,y′ N )}
wherein, M and N represent the feature points of the transformation curve segment and the curve segment to be measured;
in one embodiment, as shown in FIG. 6, an intelligent search algorithm is used to search for a segmentation interval [ R ] on the template curve 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]The step of transforming the curve segment with the maximum coincidence degree of the corresponding curve segment to be tested comprises the following steps:
s421: randomly generating an initial population P (j); j =0, for the generated individuals, judging whether the scaling a in the X direction, the scaling d in the Y direction, the translation distance l in the X direction, the translation distance m in the Y direction and the initial position offset t of the interval to be compared meet the set constraint condition, and if yes, generating the next individual; otherwise, regenerating until the constraint condition is satisfied;
s422: randomly generating a new individual in the neighborhood of the individual according to the values of the scaling a in the X direction, the scaling d in the Y direction, the translation distance l in the X direction, the translation distance m in the Y direction and the offset t of the starting position of the interval to be compared in the individual;
s423: calculating the fitness values of all parents and genetically-generated offspring, selecting S offspring according to the fitness values to form a new population P (j + 1) and replacing the population P (j);
s424: when the maximum iteration times are reached and the cycle iteration reaches the maximum generation number, stopping outputting a transformation curve segment with the maximum contact ratio with the curve segment to be detected; otherwise, go to step S422;
step S43: sampling a plurality of points at equal intervals from the transformation curve segment and the curve segment to be detected, and calculating the average distance of the points as the similar distance between the transformation curve segment and the curve segment to be detected:
wherein |. Is a distance norm, [. Sup. | ]]Represents rounding, Q is the number of sample points, F Ai F Bi Respectively a transformation curve segment and a curve segment to be tested. In other embodiments, other random intelligent search algorithms such as a particle swarm algorithm or an ant colony algorithm can be used for searching the transformation curve segment with the maximum coincidence degree with the curve segment to be detected.
Step S5: and carrying out weighted average on the similar distances corresponding to the one or more than one interval to be compared to obtain an average similar distance, and if the average similar distance is smaller than a given threshold value, judging that the three-dimensional signature track is a real signature matched with the target user information.
Specifically, the similar distances in the plurality of intervals to be compared are weighted and averaged in the following manner to obtain the similar distance between the curve to be measured and the template curve:
wherein d is i For similar distances in the respective intervals to be compared, w i And if the weights of the intervals to be compared are equal, the similar distance between the curve to be compared and the template curve is the average similar distance of all the intervals to be compared.
And giving a threshold value epsilon, if d is less than or equal to epsilon, considering that the two curves are similar, and if the three-dimensional signature track is a real signature matched with the target user information, otherwise, the three-dimensional signature track is a forged signature. The threshold epsilon may be set by user statistics. For a plurality of template curves, an average value of similar distances between the curve to be measured and each template curve may be calculated as a reference value thereof.
To solve the above technical problem, as shown in fig. 7, the present invention further provides a three-dimensional handwritten signature authentication system, including: the system comprises a body sensing device 101, a database server 102 and an authentication server 103. The motion sensing device 101 is used for collecting a three-dimensional signature track, the database server 102 is used for storing a template curve, and the authentication server 103 is used for executing the three-dimensional handwritten signature authentication method, and according to the identity ID and registration information of an operator, the submitted signature is identified and authenticated, and an authentication result is returned to the database server 102 for archiving.
The motion sensing device 101 is used for capturing user gestures through a camera to acquire a three-dimensional signature track. Or, the motion sensing device 101 is integrated on the user terminal, and is configured to acquire a three-dimensional signature trajectory by detecting a spatial motion trajectory of the user terminal.
As shown in fig. 9, the motion sensing device 101 may be an active motion sensing device, that is, an active gesture collecting device, such as a camera, a user may draw a three-dimensional signature track on a virtual plane through a gesture comparison, and the motion sensing device 101 captures a user gesture through the camera to collect the three-dimensional signature track.
In another embodiment, as shown in fig. 10, the motion sensing device 101 may be a passive motion sensing device, which may be integrated on a user terminal, and is configured to acquire a three-dimensional signature track by detecting a spatial motion track of the user terminal. For example, in fig. 10, the motion sensing device 101 may be a camera or other sensor integrated in a smart pen, the camera may capture a writing environment of a pen tip, detect a change in the writing environment through image recognition, and then reversely calculate a motion trajectory of the camera, that is, the motion trajectory of the smart pen during writing is input as a three-dimensional signature trajectory (which is not a concept with a two-dimensional handwriting of the pen tip on a paper surface), and the motion trajectory of the smart pen during writing expresses habit information of a writing posture of a user during writing.
In an embodiment of the present invention, the three-dimensional handwritten signature authentication system further includes: and the interactive display device 104 is used for inputting the identity information of the user and displaying the three-dimensional handwritten signature authentication result. The interactive display device 104 may be a computer, a handheld terminal, or a touch terminal, and is a device with display and interaction functions, and is used for an operator to perform interactive operations, including user registration, login, identity information submission, and the like.
The working process of the three-dimensional hand-written signature authentication system is as follows: the user inputs user identity information at the interactive display device 104, the authentication server 103 authenticates the user identity information and returns an authentication result to the interactive display device 104 for display, the user performs signature writing according to the display information of the interactive display device 104, when an operator hand or pen enters an induction space of the motion sensing device 101, the motion sensing device 101 starts to capture a motion track of the operator hand or pen, after the user confirms that personal signature writing is finished, the interactive device 104 can be used for signature submission, the authentication server 103 performs identification and authentication on the submitted signature track and returns the authentication result to the interactive display device 104.
The present invention also provides a computer-readable storage medium, on which a computer program for executing the above three-dimensional handwritten signature method is stored, the computer program, when executed by a processor, implementing the three-dimensional handwritten signature authentication method according to any one of the preceding claims.
The present invention may take the form of a computer program product embodied on one or more storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having program code embodied therein. Computer readable storage media, which include both non-transitory and non-transitory, removable and non-removable media, may implement any method or technology for storage of information. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of the storage medium of the computer include, but are not limited to: phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technologies, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic tape storage or other magnetic storage devices, or any other non-transmission medium, may be used to store information that may be accessed by a computing device.
The invention also provides computer equipment comprising a memory, a processor and a computer program which is stored in the memory and can be executed by the processor, wherein the processor realizes the three-dimensional handwritten signature authentication method in any item when executing the computer program.
After the three-dimensional handwritten signature authentication method, the three-dimensional handwritten signature authentication system, the storage medium and the equipment are adopted, the three-dimensional signature track of a user during signature can be detected through the somatosensory equipment, the three-dimensional signature track is different from two-dimensional handwriting in the traditional technology, the three-dimensional signature track is writing posture information determined by writing habits of the user during writing, a two-dimensional curve to be tested is obtained after the three-dimensional signature track expressing the writing posture information is rotated and transformed, then similarity comparison is carried out on the curve to be tested and a template curve corresponding to a target user, and the three-dimensional signature track is judged to be a real signature of the target user when the similarity is higher, so that the authentication result is more accurate, and the signature authentication safety is improved.
The present invention is not limited to the above-described embodiments, and various modifications and variations of the present invention are intended to be included within the scope of the claims and the equivalent technology of the present invention if they do not depart from the spirit and scope of the present invention.
Claims (10)
1. A three-dimensional handwritten signature authentication method is characterized by comprising the following steps:
acquiring a three-dimensional signature track corresponding to a three-dimensional Cartesian coordinate system data format;
respectively carrying out rotation transformation on the three-dimensional signature track along an X axis and a Y axis of an XYZ reference system, selecting a target X axis rotation angle and a target Y axis rotation angle when the projections of the three-dimensional signature track on an XZ plane and a YZ plane are minimum in the rotation transformation process of the three-dimensional signature track, and obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured when the three-dimensional signature track is rotated to the target X axis rotation angle and the target Y axis rotation angle;
acquiring a template curve corresponding to the target user information;
segmenting the template curve and the curve to be compared, wherein each segment of the template curve has a corresponding interval to be compared on the curve to be compared; carrying out similarity transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and calculating the average distance between the curve to be tested in the interval to be compared and the transformation curve segment as the similarity distance;
and carrying out weighted average on the similar distances corresponding to the one or more than one interval to be compared to obtain an average similar distance, and if the average similar distance is smaller than a given threshold value, judging that the three-dimensional signature track is a real signature matched with the target user information.
2. The method for authenticating a three-dimensional handwritten signature according to claim 1, wherein the step of performing rotation transformation on the three-dimensional signature trajectory along X-axis and Y-axis of XYZ reference system, respectively, and in the rotation transformation of the three-dimensional signature trajectory, selecting a target X-axis rotation angle and a target Y-axis rotation angle at which the projection of the three-dimensional signature trajectory on XZ and YZ planes is minimum, and obtaining a projection curve of the three-dimensional signature trajectory on XY plane as a curve to be measured when the three-dimensional signature trajectory is rotated to the target X-axis rotation angle and the target Y-axis rotation angle includes:
step S201: carrying out X-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein the rotation transformation formula is as follows:
theta is the rotation angle of the X-axis rotation transformation and belongs to the angle range of [ -90, 90 ];
step S202: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each X-axis rotation transformation, and selecting the angle theta corresponding to the one-time rotation transformation with the minimum variance as a target X-axis rotation angle;
step S203: carrying out Y-axis rotation transformation on the three-dimensional signature track for a plurality of times at set angle intervals, wherein the rotation transformation formula is as follows:
step S204: calculating the variance of the Z coordinate of the sampling point on the three-dimensional signature track after each Y-axis rotation transformation, and selecting the angle corresponding to the rotation transformation with the minimum variance
As a target Y-axis rotation angle;
step S205: taking the value of theta as the target X-axis rotation angle,
when the value is taken as the rotation angle of the target Y axis, the transformation matrix T = T 1 T 2 And obtaining a projection curve of the three-dimensional signature track on an XY plane as a curve to be measured.
3. The method according to claim 2, wherein the step of obtaining the template curve corresponding to the target user information comprises:
acquiring a template three-dimensional signature track corresponding to a three-dimensional Cartesian coordinate coefficient data format;
and executing the steps S201-S205 on the template three-dimensional signature track to obtain a projection curve of the template three-dimensional signature track on an XY plane as a template curve.
4. The method for authenticating a three-dimensional handwritten signature as claimed in claim 1, wherein the step of performing a similar transformation on the template curve segment in the interval to be compared to obtain a transformation curve segment corresponding to the interval to be compared and having the maximum coincidence degree of the curve segment to be tested, and the step of calculating the average distance between the curve to be tested in the interval to be compared and the transformation curve segment as the similar distance comprises:
for a section [ C ] to be compared with an initial position offset of t 1 ,C 2 ]Carrying out similarity transformation on the template curve according to the transformation matrix T;
the transformation matrix T is
a. d is scaling in X and Y directions respectively; l and m are translation distances in X and Y directions respectively;
searching by using an intelligent search algorithm to obtain a template curve segmentation interval R 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]The corresponding transformation curve section F with the maximum contact ratio of the curve section to be measured A The template curve segment corresponds to a transformation curve segment F A Comprises the following steps:
F A ＝{(x 1 ,y 1 ),…(x 2 ,y 2 ),…(x M ,y M )}
the interval [ C ] to be compared 1 ,C 2 ]Corresponding curve segment F to be measured B Comprises the following steps:
F B ＝{(x′ 1 ,y′ 1 ),…(x′ 2 ,y′ 2 ),…(x′ N ,y′ N )}
wherein M and N represent a transformation curve segment F A And the curve segment F to be measured B The number of feature points;
sampling a plurality of sampling points at equal intervals from the transformation curve segment and the curve segment to be detected, and calculating the average distance between the sampling points at the corresponding positions as the similar distance between the transformation curve segment and the curve segment to be detected:
wherein |. Is a distance norm, [. Sup. | ]]Represents rounding, Q is the number of sample points, F Ai F Bi Respectively a transformation curve segment and a sampling point of a curve segment to be detected.
5. The method according to claim 4, wherein the template curve segment [ R ] is obtained by searching through an intelligent search algorithm 1 ,R 2 ]And the interval [ C ] to be compared 1 ,C 2 ]The corresponding step of transforming the curve segment with the maximum contact ratio of the curve segment to be tested comprises the following steps:
s421: randomly generating an initial population P (j); wherein j =0;
s422: randomly generating a new individual in the neighborhood of the individual according to the values of the scaling a in the X direction, the scaling d in the Y direction, the translation distance l in the X direction, the translation distance m in the Y direction and the initial position offset t of the interval to be compared in the individual;
s423: calculating the fitness values of all parents and genetically-generated offspring, selecting S offspring according to the fitness values to form a new population P (j + 1) and replacing the population P (j);
s424: when the maximum iteration times are reached and the loop iteration reaches the maximum generation number, stopping, and selecting the transformation curve segment with the maximum coincidence degree with the curve segment to be detected for output; otherwise, go to step S422.
6. The method according to claim 1, wherein the step of obtaining the template curve corresponding to the target user information comprises:
and receiving and authenticating the input target user information, and acquiring a template curve corresponding to the target user information in the server after the authentication is passed.
7. A three-dimensional handwritten signature authentication system, comprising: the system comprises a body sensing device, a database server and an authentication server; the motion sensing device is used for collecting a three-dimensional signature track, the database server is used for storing the template curve, and the authentication server is used for executing the three-dimensional handwritten signature authentication method of any one of claims 1 to 6;
the motion sensing equipment is used for capturing user gestures through a camera and acquiring a three-dimensional signature track; or,
the motion sensing device is integrated on the user terminal and used for acquiring a three-dimensional signature track by detecting a spatial motion track of the user terminal.
8. The three-dimensional handwritten signature authentication system according to claim 7, further comprising: and the interactive display equipment is used for receiving the input target user information and displaying the three-dimensional handwritten signature authentication result.
9. A computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method for authenticating a three-dimensional handwritten signature according to any one of claims 1 to 6.
10. A computer arrangement comprising a memory, a processor and a computer program stored in the memory and executable by the processor, the processor implementing the steps of the method of authenticating a three-dimensional handwritten signature according to any of claims 1-6 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911213154.8A CN110942042B (en) | 2019-12-02 | 2019-12-02 | Three-dimensional handwritten signature authentication method, system, storage medium and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911213154.8A CN110942042B (en) | 2019-12-02 | 2019-12-02 | Three-dimensional handwritten signature authentication method, system, storage medium and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110942042A CN110942042A (en) | 2020-03-31 |
| CN110942042B true CN110942042B (en) | 2022-11-08 |
Family
ID=69908680
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911213154.8A Active CN110942042B (en) | 2019-12-02 | 2019-12-02 | Three-dimensional handwritten signature authentication method, system, storage medium and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110942042B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115146248A (en) * | 2021-03-31 | 2022-10-04 | 中国电信股份有限公司 | Verification method, device, electronic device and storage medium |
| CN119027858B (en) * | 2024-10-18 | 2024-12-27 | 广东微模式软件股份有限公司 | True and false signature action detection method and device for double-record quality inspection |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008046781A (en) * | 2006-08-11 | 2008-02-28 | National Institute Of Advanced Industrial & Technology | Handwritten signature personal authentication system based on time series of contact force |
| CN105912910A (en) * | 2016-04-21 | 2016-08-31 | 武汉理工大学 | Cellphone sensing based online signature identity authentication method and system |
| CN107657241A (en) * | 2017-10-09 | 2018-02-02 | 河海大学常州校区 | A kind of signature true or false identification system towards signature pen |
| CN108139856A (en) * | 2016-08-31 | 2018-06-08 | 华为技术有限公司 | Signature authentication method, terminal, writing pencil and system |
| CN108536314A (en) * | 2017-03-06 | 2018-09-14 | 华为技术有限公司 | Method for identifying ID and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI569176B (en) * | 2015-01-16 | 2017-02-01 | 新普科技股份有限公司 | Method and system for identifying handwriting track |
-
2019
- 2019-12-02 CN CN201911213154.8A patent/CN110942042B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008046781A (en) * | 2006-08-11 | 2008-02-28 | National Institute Of Advanced Industrial & Technology | Handwritten signature personal authentication system based on time series of contact force |
| CN105912910A (en) * | 2016-04-21 | 2016-08-31 | 武汉理工大学 | Cellphone sensing based online signature identity authentication method and system |
| CN108139856A (en) * | 2016-08-31 | 2018-06-08 | 华为技术有限公司 | Signature authentication method, terminal, writing pencil and system |
| CN108536314A (en) * | 2017-03-06 | 2018-09-14 | 华为技术有限公司 | Method for identifying ID and device |
| CN107657241A (en) * | 2017-10-09 | 2018-02-02 | 河海大学常州校区 | A kind of signature true or false identification system towards signature pen |
Non-Patent Citations (3)
| Title |
|---|
| "Online Signature Verification Based on a Single Template via Elastic Curve Matching";Huacheng Hu等;《Sensors》;20191107;第1-23页 * |
| "基于曲率的在线签名认证研究";谭华;《中国优秀硕士学位论文全文数据库 信息科技辑》;20190215;I136-474 * |
| 基于空间曲线弹性匹配的在线手写签名鉴别;金涌等;《华中理工大学学报》;19990530(第05期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110942042A (en) | 2020-03-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Fingerprint verification using spectral minutiae representations | |
| Sun et al. | Ordinal palmprint represention for personal identification [represention read representation] | |
| Drost et al. | Model globally, match locally: Efficient and robust 3D object recognition | |
| Jain et al. | Exploring orientation and accelerometer sensor data for personal authentication in smartphones using touchscreen gestures | |
| Das et al. | Recent advances in biometric technology for mobile devices | |
| CN100452084C (en) | Image recognition apparatus and its method | |
| Sarkar et al. | Deep feature-based face detection on mobile devices | |
| CN103595538B (en) | Identity verification method based on mobile phone acceleration sensor | |
| CN103593673B (en) | A kind of on-line trial authentication method based on dynamic threshold | |
| Islam et al. | A review of recent advances in 3D ear-and expression-invariant face biometrics | |
| Nedjah et al. | Efficient fingerprint matching on smart cards for high security and privacy in smart systems | |
| CN107657241A (en) | A kind of signature true or false identification system towards signature pen | |
| Soltanpour et al. | Multimodal 2D–3D face recognition using local descriptors: pyramidal shape map and structural context | |
| Mohammed et al. | State-of-the-art in handwritten signature verification system | |
| CN110942042B (en) | Three-dimensional handwritten signature authentication method, system, storage medium and equipment | |
| JP2014182754A (en) | Learning device, learning method, and program | |
| Badrinath et al. | Score level fusion of voting strategy of geometric hashing and SURF for an efficient palmprint-based identification | |
| Elisha Raju et al. | A two-level security system based on multimodal biometrics and modified fusion technique | |
| CN106951751A (en) | A kind of sensor-based smart mobile phone unlocking recognition methods | |
| Lu et al. | SCUT-MMSIG: a multimodal online signature database | |
| CN209486697U (en) | A posture recognition authentication device based on deep learning | |
| CN118334756A (en) | Electronic signature handwriting source classification method and system based on bimodal feature fusion | |
| Khursheed et al. | AR model based human identification using ear biometrics | |
| Al-Jarrah et al. | Finger-drawn signature verification on touch devices using statistical anomaly detectors | |
| Ganapathi et al. | Secure multimodal access with 2D and 3D ears |
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 |