CN110135145A - Click identifying code method based on turing test - Google Patents
Click identifying code method based on turing test Download PDFInfo
- Publication number
- CN110135145A CN110135145A CN201910431895.7A CN201910431895A CN110135145A CN 110135145 A CN110135145 A CN 110135145A CN 201910431895 A CN201910431895 A CN 201910431895A CN 110135145 A CN110135145 A CN 110135145A
- Authority
- CN
- China
- Prior art keywords
- identifying code
- signal language
- background
- layer
- random
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06F21/36—User authentication by graphic or iconic representation
-
- 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
- G06F21/44—Program or device authentication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2221/00—Indexing scheme relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F2221/21—Indexing scheme relating to G06F21/00 and subgroups addressing additional information or applications relating to security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F2221/2133—Verifying human interaction, e.g., Captcha
Landscapes
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The present invention relates to the click identifying code methods based on turing test, comprising: A. generates Background From Layer and paints at random, forms identifying code display area；B. identifying code figure is generated at random, forms identifying code by covering of part Background From Layer, identifying code is shown in identifying code viewing area；C. signal language background is generated；D. signal language figure layer and random signal language picture are generated, language background, signal language figure layer and signal language picture is will be prompted to and is bonded to form signal language, the mark with number attribute is included in signal language；E. Background From Layer is marked according to the display position of identifying code, further according to the number attribute in the corresponding movement of the position of label addition；F. the number attribute of user's mark according to signal language is clicked corresponding number to identifying code and is verified.The difficulty that identifying code is cracked is greatly improved in the present invention, while significantly simplifying the operation of user, while can also allow the accessible use of dyschromatopsia personage.
Description
Technical field
The present invention relates to the verification methods of electronic information, are concretely the click identifying code methods based on turing test.
Background technique
Identifying code (CAPTCHA) is one kind of turing test, and user is computer or people for identification.Identifying code is main
For preventing webpage or system by the explosion of malice password, pouring water, brush ticket etc.；Identifying code can effectively prevent hacker and carry out not to system
User information is stolen in disconnected login attempt.The usual manner of many systems using identifying code, for example, by system generate a string with
Character, number or the picture that machine generates identify verification code information by user, and validation information submits website authentication, verifying at
System function could be used after function.
Traditional identifying code generallys use graphic verification, answer verifying, sliding block verifying, clicks the modes such as verifying, tradition side
The identifying code of formula can satisfy the primary demand of verifying, but with the fast development of Internet technology, artificial intelligence and hacker's skill
The continuous promotion of art, traditional identifying code cannot fully meet current operation demand and turing test standard.
1, graphical verification code implementation: as shown in Figure 1, server-side generates a picture for doing identifying code back first
Scape, by fill assembly and plus miscellaneous component processing is filled to background picture, then generated at random by identifying code formation component specific
The verifying number of digit, character or text, distort identifying code, are rotated, deforms etc. and handled by deformed component, then will test
Card code and background carry out fitting and are shown, and carry out input verification operation by page prompts, complete prompt operation then verifying at
Function, otherwise authentication failed is verified again.
The verification mode is with the presence of following problem: 1, safety issue: the Photoshop and identifying code figure layer of background picture
The difference of biggish contrast, brightness, pixel etc. is easy to separate background and identifying code by specific program, to crack
Identifying code provides condition.Because the identifying code number generated at random, character, text identifying code have only done distortion, rotation, deformation
Processing, the pattern of identifying code is limited, hacker can by picture cutting technique by identifying code be divided into it is independent number, character or
Text identifies that identifying code completion cracks by picture recognition technology, or carrying out modeling study by convolutional neural networks will
All identifying code patterns generated are recorded and are compared completion and cracked.2, experience problem: the design is generally using input verifying
Mode fail to fully take into account now widely used mobile terminal mode and scene, mobile terminal are limited to touch screen,
Verifying code character is manually entered and increases the verification time, reduces the experience effect of user.3, accessible to use problem: the design
By the way of color background and colored identifying code fitting, which needs to have identifying code and background colour higher discrimination energy
Power, but the use of the dyschromatopsia personage in the whole world more than 200,000,000 is not considered in the design, and dyschromatope is caused not know correctly
Other identifying code.
2, answer identifying code implementation: as shown in Fig. 2, server-side generates a picture for doing identifying code back first
Scape by fill assembly and adds miscellaneous component to be filled processing to background picture, then passes through special algorithm by identifying code formation component
Specific topic identifying code is generated at random, the processing such as is distorted by deformed component, rotate, deform to identifying code, then will test
Card code and background carry out fitting and are shown, and user carries out answer input according to identifying code prompt information, test if answer correctly
It demonstrate,proves successfully, otherwise authentication failed is verified again.
The verification mode has following problem: other than the drawback identical as above-mentioned graphical verification code implementation, also having
Have crowd's confinement problems: to verifier, there are certain requested knowledges on a certain field for the design, do not have world civilization, contracting
It is small to use crowd's range.
3, sliding block identifying code implementation: sliding block identifying code is by sliding with background picture, the picture of the scarce block of filling for lacking block
Block, one can drag slider button composition, follow sliding by dragging slider bar slider picture-puzzle, slidable puzzle to correct position simultaneously
Unclamp finger, it may appear that be proved to be successful information, sliding into errors present will appear authentication failed information.
The verification mode is with the presence of following problem: 1, safety issue: shade at the scarce block of background picture, the pixel at this
For pixel with whole picture there are apparent difference, automated procedures can pass through pixel difference comparison technology, it is easy to know
Not Chu sliding block need the positional that drags to, then sliding block is moved to completion at verifying by capturing slider button and is cracked.2,
The accessible problem that uses: by the way of the design is bonded using color background and colored identifying code, which need to identifying code with
Background colour has higher resolving ability, but the use of the dyschromatopsia personage in the whole world more than 200,000,000 is not considered in the design, leads to color
Feel that obstacle person can not correctly identify identifying code.
4, click identifying code implementation: first server-side store a verification information library, the information bank include picture or
Text or figure etc., identifying code formation component recall verification information out of library at random, then by adding miscellaneous component to transferring at random
Verification information is distorted, adds interfering line, the modes such as variegated is added to be handled, and server-side generates an identifying code prompt random
Language, user clicks according to identifying code signal language selects correct verification information to complete verifying, and otherwise authentication failed is verified again.
The verification mode has following problem: 1, safety issue: the verification information that verification information inventory is put is limited, attacker
Can be learnt by CNN convolutional neural networks to verification informations all in library, the identifying code that crawl generates at random in verifying into
Row identification, then grabs the identifying code signal language generated at random and is identified, according to signal language require to the identifying code recognized by
Sequence, which is clicked, to be completed to crack.2, the accessible problem that uses: the design, should by the way of color background and colored identifying code fitting
The resolving ability that mode needs that it is certain to have identifying code and background colour, the use for dyschromatopsia personage can have a foregone conclusion
It is sex-limited.
Summary of the invention
Aiming at the problems existing in the prior art, the present invention provides a kind of click identifying code side based on turing test
Method is not only able to improve the difficulty that is cracked of identifying code, and can also simplify the operation of user, while can also allow dyschromatopsia
The accessible use of personage.
The present invention is based on the click identifying code methods of turing test, comprising steps of
A. a Background From Layer is generated, is painted at random to Background From Layer, identifying code display area is formed；
B. identifying code figure is generated at random, then intercepts illiteracy of the Background From Layer described in part as the identifying code figure
Skin shows identifying code in the identifying code viewing area using the identifying code figure after covering as identifying code；
C. signal language background is generated；
D. signal language figure layer and random signal language picture are generated, will be prompted to language background, signal language figure layer and prompt sonagram
Piece is successively bonded, and forms signal language, and 1 mark with number attribute is included at least in signal language；
E. the Background From Layer is marked according to the display position of identifying code, further according to identifying institute described in signal language
The number attribute of expression, the corresponding movement of addition in the position of the label；
F. the number attribute of user's mark according to signal language is clicked corresponding number to the identifying code and is carried out
Verifying.
Compared with conventional authentication code, most significant difference, which is verification process not, to be carried out to identifying code method of the invention
Intuitive operation or logic judgment do not need the stock of knowledge that user has certain field yet, but according to the number in signal language
Mark carries out the click of corresponding number to identifying code, and identification to number mark and to the click of identifying code can only by people come
Realize, though by the technologies such as hacker will be prompted to sonagram piece and signal language figure layer separation, can not also know wherein number identify
Content, and virtual click is carried out to identifying code.And by the way of the corresponding number of identifying code that adopting consecutive click chemical reaction generates at random
Also it solves the problems, such as the experience in mobile phone terminal, the time that system is verified has been shortened in 5ms, no matter in the end PC or mobile phone
End all need to only carry out adopting consecutive click chemical reaction operation, remove user from and typewrite the problems such as inconvenient.
Further, it is only painted at random with black and white in step A to Background From Layer.By black to Background From Layer
White colouring solves the problems, such as that dyschromatopsia personage uses color detection when identifying code.
Further, step D first judges corresponding to the access data of active user when generating random signal language picture
Security level, corresponding signal language picture is randomly selected in server-side dictionary library further according to the security level judged, together
When server-side dictionary library be periodically updated.The security level of user is higher, the prompt randomly selected in server-side dictionary library
The content of sonagram piece is simpler, on the contrary then content is more complicated, has thus reached simplified high security user's operation, reinforces lower security
The purpose of property operation demonstration.
Specifically, including: IP address to active user, machine to the judgement of security level corresponding to active user
Code, login times, refreshing frequency, authentication failed number, at least one for being proved to be successful number and verifying in frequency are used as data
Source is collected to obtain behavioral data j, and machine learning model obtains the security level of active user after analyzing behavioral data j,
The security level of obtained active user is obtained to the safety regulation of active user afterwards compared with the standard class m of definition；So
Afterwards further according to the safety regulation of active user according to randomly selecting corresponding signal language picture in server-side dictionary library.
Further, before step E further include: will be prompted to language or the mark with number attribute generates audio；In step F
User clicks corresponding number to the identifying code and verifies according to the digital information in the audio.Signal language or tool
Having the mark of number attribute can not only be embodied by visual effect, can also be solved in this way by way of voice prompting
The confinement problems that the signal language of part special population obtains, for the user of signal language acquisition of information difficulty can pass through language
Sound prompt is made corresponding clicking operation and is verified.
On this basis, step A includes:
A1. the blank background z that a length of x wide is y is generated；
A2. pattern c is obtained after arbitrarily generating pattern a and pattern b, pattern a and pattern b mixing, pattern c is carried out on random
Color fills to obtain pattern c1；
A3., pattern c1 after colouring is filled into the background generated in the blank background z as identifying code display area
z1。
Preferably, the identifying code figure that step B is generated at random is 3-D graphic.The three-dimensional graph generated at random is made
For identifying code, and using the random Background From Layer intercepted at random as the covering of 3 dimensional drawing, this identifying code for
It is the random pixel of a pile for automatic identification program, and tells two-dimensional background and three for what can visually be apparent
Identifying code is tieed up, for machine learning captures program, because identifying code itself is that there is no shape library, machine learning
Capture program can not construct huge condition code library and be cracked.
Preferably, in step C, signal language Background From Layer is first generated, then the signal language Background From Layer is carried out plus done
It disturbs line and/or forms the signal language background after adding interference processes pixel.
Preferably, at at least one for the signal language picture generated at random first being deformed, being distorted or being rotated in step D
It after reason, then will be prompted to sonagram layer and signal language picture is bonded, form signal language.
By adding miscellaneous treated signal language background and signal language that can further increase signal language by the difficulty of Brute Force
Degree, is difficult to automated procedures more.
Beneficial effect the present invention is based on the click identifying code method of turing test includes: 1, uses machine learning model
Come intelligent decision people and automated procedures, constructs powerful security rule base, motivation in parallel according to Skinner theory is learned at heart
Device learning model to handle normal authentication behavior and the blasting behavior of automated procedures of people in real time.2, by generate at random three
Solid figure is tieed up as the random background patterns of the background patterns of identifying code and Stochastic colored generation, this identifying code for
Almost can not identify and crack for automatic identification program, and naked eyes can be apparent tell two-dimensional background and
Three-dimensional verification code.3, optimize user experience: being solved by way of the corresponding number of the identifying code that adopting consecutive click chemical reaction generates at random
Mobile phone terminal experiences problem, and the time of verifying is shortened in 5ms, no matter at the end PC or mobile phone terminal all need to only carry out continuity point
Operation is hit, user is eliminated and typewrites the problems such as inconvenient.4, it solves the problems, such as accessible use: using chequered with black and white combination
Solve the problems, such as that dyschromatopsia personage uses color detection when identifying code to identifying code and Background From Layer.5, crowd office is solved
Sex-limited problem: the signal language that part special population is solved in such a way that the visual effect of signal language is shown with voice prompting obtains
Confinement problems, for signal language acquisition of information difficulty user can be made by voice prompting corresponding clicking operation into
Row verifying.
Specific embodiment with reference to embodiments is described in further detail above content of the invention again.
But the range that this should not be interpreted as to the above-mentioned theme of the present invention is only limitted to example below.Think not departing from the above-mentioned technology of the present invention
In the case of thinking, the various replacements or change made according to ordinary skill knowledge and customary means should all be included in this hair
In bright range.
Detailed description of the invention
Fig. 1 is a kind of verifying form of existing graphical verification code.
Fig. 2 is a kind of verifying form of existing answer identifying code.
Fig. 3 is a kind of verifying form of existing sliding block identifying code.
Fig. 4 is a kind of existing verifying form for clicking identifying code.
Fig. 5 is the flow chart of the click identifying code method the present invention is based on turing test.
Specific embodiment
The present invention is based on the click identifying code methods of turing test as shown in Figure 5, comprising steps of
A. a Background From Layer is generated, is only painted at random with black and white to Background From Layer, it is aobvious to form identifying code
Show region:
A1. the blank background z that a length of x wide is y is generated；
A2. obtain pattern c after arbitrarily generating pattern a and pattern b, pattern a and pattern b mixing, to pattern c only with black and
White carries out random colouring filling and obtains pattern c1；
A3. the pattern c1 after black and white being painted is filled into the blank background z and generates as identifying code display area
Background z1.
B. the identifying code figure of 3-D graphic is generated at random, and the area of the 3-D graphic is less than identifying code display area
Area, cutting can be placed in identifying code display area.Background From Layer described in part is intercepted again to test as the 3-D graphic
Identifying code is shown and is shown in the identifying code using the identifying code figure after covering as identifying code by the covering of card code figure
In area；
C. signal language Background From Layer is generated, then the signal language Background From Layer is carried out adding interfering line and/or adds interference
Signal language background is formed after processes pixel；
D1. signal language figure layer and random signal language picture are generated, and will be prompted to sonagram layer and signal language picture is become
After shape, distortion or rotation, it will be prompted to language background, signal language figure layer and signal language picture and be successively bonded, form signal language, prompting
1 mark with number attribute is included at least in language.Such as signal language can be the Chinese idiom with number, or have number
A word, several words etc..Specific steps are as follows:
When generating random signal language picture, security level corresponding to the access data of active user is first judged, wrap
It includes the IP address to active user, machine code, login times, refreshing frequency, authentication failed number, be proved to be successful number and verifying
At least one in frequency is collected to obtain behavioral data j as data source, after machine learning model analyzes behavioral data j
The security level of active user is obtained, the security level of obtained active user is obtained afterwards compared with the standard class m of definition
To the safety regulation of active user.The operation of weight plus/minus is carried out according to security level of the corresponding safety regulation to active user,
The higher validation difficulty of weight is smaller, and the content of signal language picture is simpler；Conversely, the lower validation difficulty of weight is bigger, signal language
The content of picture is more complicated.Corresponding prompt is randomly selected in server-side dictionary library according to the result after the operation of weight plus/minus
Sonagram piece, while server-side dictionary library is periodically updated.
For example, the verifier of grade 1 is determined as that high-quality user, the behavioral data j of this kind of user are above machine learning
The standard class m that modeler defines can trigger validation difficulty low authentication mechanism when this kind of user's checking.Once the behavior of user
Data are unsatisfactory for when standard class m validation difficulty then automatic lifting；The verifier of grade 2 is determined as ordinary user, such verifying
The behavioral data j of person complies with standard grade m, such verifier can trigger mechanism to be confirmed, needs the behavior by a period of time t
Data k is learnt to be classified again；The verifier of grade 3 is determined as malicious user, the behavioral data j of such verifier is lower than
Standard class m, such verifier can trigger the authentication mechanism of validation difficulty promotion.
D2. system will also prompt language or the mark with number attribute generates audio.
E. the Background From Layer is marked according to the display position of identifying code, further according to identifying institute described in signal language
The number attribute of expression, the corresponding movement of addition in the position of the label；
F. when user's checking, system calls corresponding safety regulation by obtaining the different brackets of active user, and to working as
The grade of preceding user carries out weight plus-minus operation, and the higher validation difficulty of weight is smaller, and authentication mechanism is simpler, otherwise weight is lower
Validation difficulty is bigger, and authentication mechanism is more complicated.User is according to the number attribute identified in signal language or audio, to the verifying
Code-point hits corresponding number and is verified.
Claims (9)
1. the click identifying code method based on turing test, feature comprising steps of
A. a Background From Layer is generated, is painted at random to Background From Layer, identifying code display area is formed；
B. identifying code figure is generated at random, then intercepts covering of the Background From Layer described in part as the identifying code figure, it will
Identifying code figure after covering shows identifying code in the identifying code viewing area as identifying code；
C. signal language background is generated；
D. signal language figure layer and random signal language picture are generated, will be prompted to language background, signal language figure layer and signal language picture according to
Secondary fitting forms signal language, and 1 mark with number attribute is included at least in signal language；
E. the Background From Layer is marked according to the display position of identifying code, further according to represented by mark described in signal language
Number attribute, the corresponding movement of addition in the position of the label；
F. the number attribute of user's mark according to signal language is clicked corresponding number to the identifying code and is tested
Card.
2. the click identifying code method based on turing test as described in claim 1, it is characterized in that: to Background in step A
Layer is only painted with black and white at random.
3. the click identifying code method based on turing test as described in claim 1, it is characterized in that: step D is being generated at random
Signal language picture when, first judge security level corresponding to the access data of active user, further according to judged safety etc.
Grade randomly selects corresponding signal language picture in server-side dictionary library, while server-side dictionary library is periodically updated.
4. the click identifying code method based on turing test as claimed in claim 3, it is characterized in that: to corresponding to active user
The judgement of security level include: the IP address to active user, machine code, login times, refreshing frequency, authentication failed time
Number, at least one for being proved to be successful number and verifying in frequency are collected to obtain behavioral data j, machine learning as data source
Model obtains the security level of active user after analyzing behavioral data j, by the security level and definition of obtained active user
Standard class m the safety regulation for obtaining active user more afterwards；Then further according to the safety regulation of active user according to from
Corresponding signal language picture is randomly selected in server-side dictionary library.
5. the click identifying code method based on turing test as described in claim 1, it is characterized in that: before step E further include:
It will be prompted to language or the mark with number attribute generate audio；User is according to the digital information in the audio in step F, to institute
The identifying code stated is clicked corresponding number and is verified.
6. the click identifying code method based on turing test as described in one of claim 1 to 5, it is characterized in that: step A packet
It includes:
A1. the blank background z that a length of x wide is y is generated；
A2. pattern c is obtained after arbitrarily generating pattern a and pattern b, pattern a and pattern b mixing, random colouring is carried out to pattern c and is filled out
It fills to obtain pattern c1；
A3., pattern c1 after colouring is filled into the background z1 generated in the blank background z as identifying code display area.
7. the click identifying code method based on turing test as described in one of claim 1 to 5, it is characterized in that: step B is random
The identifying code figure of generation is 3-D graphic.
8. the click identifying code method based on turing test as described in one of claim 1 to 5, it is characterized in that: in step C,
Signal language Background From Layer is first generated, then the signal language Background From Layer is carried out adding interfering line and/or adds interference processes pixel
The signal language background is formed afterwards.
9. the click identifying code method based on turing test as described in one of claim 1 to 5, it is characterized in that: it is right in step D
After at least one processing that the signal language picture generated at random is first deformed, distorted or rotated, then it will be prompted to sonagram layer and mention
Show that sonagram piece is bonded, forms signal language.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910431895.7A CN110135145A (en) | 2019-05-23 | 2019-05-23 | Click identifying code method based on turing test |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910431895.7A CN110135145A (en) | 2019-05-23 | 2019-05-23 | Click identifying code method based on turing test |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110135145A true CN110135145A (en) | 2019-08-16 |
Family
ID=67572660
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910431895.7A Pending CN110135145A (en) | 2019-05-23 | 2019-05-23 | Click identifying code method based on turing test |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110135145A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111563941A (en) * | 2020-05-08 | 2020-08-21 | 拉扎斯网络科技（上海）有限公司 | Picture verification code generation method, verification method, device and related equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971045A (en) * | 2013-01-25 | 2014-08-06 | 苏州精易会信息技术有限公司 | Click type verification code implementation method |
| CN105184130A (en) * | 2015-09-28 | 2015-12-23 | 北京橙鑫数据科技有限公司 | Terminal unlocking method and device |
| CN105577367A (en) * | 2015-12-22 | 2016-05-11 | 安徽一拓通信科技集团股份有限公司 | Method for generating and verifying verification code |
| CN107104973A (en) * | 2017-05-09 | 2017-08-29 | 北京潘达互娱科技有限公司 | The method of calibration and device of user behavior |
-
2019
- 2019-05-23 CN CN201910431895.7A patent/CN110135145A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971045A (en) * | 2013-01-25 | 2014-08-06 | 苏州精易会信息技术有限公司 | Click type verification code implementation method |
| CN105184130A (en) * | 2015-09-28 | 2015-12-23 | 北京橙鑫数据科技有限公司 | Terminal unlocking method and device |
| CN105577367A (en) * | 2015-12-22 | 2016-05-11 | 安徽一拓通信科技集团股份有限公司 | Method for generating and verifying verification code |
| CN107104973A (en) * | 2017-05-09 | 2017-08-29 | 北京潘达互娱科技有限公司 | The method of calibration and device of user behavior |
Non-Patent Citations (1)
| Title |
|---|
| 李霞 等: ""UI无障碍色彩交互的认知设计原则"", 《UI交互色彩设计》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111563941A (en) * | 2020-05-08 | 2020-08-21 | 拉扎斯网络科技（上海）有限公司 | Picture verification code generation method, verification method, device and related equipment |
| CN111563941B (en) * | 2020-05-08 | 2023-12-22 | 拉扎斯网络科技（上海）有限公司 | Picture verification code generation method, verification method, device and related equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Invisible backdoor attacks on deep neural networks via steganography and regularization | |
| CN108229130B (en) | Verification method and device | |
| Serwadda et al. | When kids' toys breach mobile phone security | |
| Bursztein et al. | Easy does it: More usable CAPTCHAs | |
| US10387645B2 (en) | Method for recognizing if a user of an electronic terminal is a human or a robot | |
| US8990959B2 (en) | Manipulable human interactive proofs | |
| Cha et al. | Boosting the guessing attack performance on android lock patterns with smudge attacks | |
| CN108650226B (en) | A kind of login validation method, device, terminal device and storage medium | |
| CN104408341A (en) | Smart phone user identity authentication method based on gyroscope behavior characteristics | |
| CN109902459B (en) | Man-machine identification verification method, system, equipment and storage medium for WEB page | |
| CN105323073A (en) | Identity authentication method, identity authentication device and identity authentication system | |
| CN112883359B (en) | Dynamic verification method and system | |
| CN106778151A (en) | Method for identifying ID and device based on person's handwriting | |
| CN106250756A (en) | Generation method, verification method and the related device of identifying code | |
| TWI770947B (en) | Verification method and verification apparatus based on attacking image style transfer | |
| CN111143812B (en) | Login authentication method based on graphics | |
| Dinh et al. | Recent advances of Captcha security analysis: a short literature review | |
| CN109543374B (en) | Sliding type identifying code for resisting deep neural network identification | |
| CN107358088A (en) | Verification method and system based on clock | |
| CN107742068A (en) | A kind of implicit identity authorization system of the multi-source of smart machine and method | |
| KR101363668B1 (en) | Apparatus and method for authentication user using captcha | |
| Dinh et al. | Human-artificial intelligence approaches for secure analysis in CAPTCHA codes | |
| CN110135145A (en) | Click identifying code method based on turing test | |
| CN111353140A (en) | Verification code generation and display method, device and system | |
| CN110263530B (en) | Authentication method and device for password reset request |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190816 |
|
| RJ01 | Rejection of invention patent application after publication |