CN106384595B - A kind of payment platform login method and device based on speech cipher - Google Patents
A kind of payment platform login method and device based on speech cipher Download PDFInfo
- Publication number
- CN106384595B CN106384595B CN201610703600.3A CN201610703600A CN106384595B CN 106384595 B CN106384595 B CN 106384595B CN 201610703600 A CN201610703600 A CN 201610703600A CN 106384595 B CN106384595 B CN 106384595B
- Authority
- CN
- China
- Prior art keywords
- audio signal
- model
- characteristic point
- frame
- point
- 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 51
- 230000005236 sound signal Effects 0.000 claims abstract description 108
- 230000008569 process Effects 0.000 claims description 26
- 238000012545 processing Methods 0.000 claims description 22
- 238000012544 monitoring process Methods 0.000 claims description 18
- 238000009432 framing Methods 0.000 claims description 17
- 238000012216 screening Methods 0.000 claims description 13
- 238000005070 sampling Methods 0.000 claims description 9
- 238000007667 floating Methods 0.000 claims description 7
- 238000001228 spectrum Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 2
- 239000000284 extract Substances 0.000 claims 2
- 238000010586 diagram Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
- G10L17/22—Interactive procedures; Man-machine interfaces
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
- G10L17/02—Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
- G10L17/06—Decision making techniques; Pattern matching strategies
- G10L17/08—Use of distortion metrics or a particular distance between probe pattern and reference templates
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/32—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials
- H04L9/3226—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials using a predetermined code, e.g. password, passphrase or PIN
- H04L9/3231—Biological data, e.g. fingerprint, voice or retina
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Security & Cryptography (AREA)
- Biodiversity & Conservation Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Telephone Function (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention discloses a kind of payment platform login method and device based on speech cipher, account and audio signal including receiving user's input, the audio signal is decomposed, establishes characteristic point to model to each frame;It is inquired in preset user message table, obtains audio signal of the characteristic point to model for including one or more foundation;The characteristic point of each audio signal found matches model with newly-established all characteristic points to model, the corresponding account information of audio signal of successful match is obtained in the user message table if successful match, judge whether corresponding account information is identical as the account information of user's input, payment platform is then logged in if they are the same, the login failure if not identical;The direct login failure if matching is unsuccessful.Therefore, the payment platform login method and device based on speech cipher solves the problems, such as that existing password safety is poor, not convenient and quick enough.
Description
Technical field
The present invention relates to fields of communication technology, particularly relate to a kind of payment platform login method and dress based on speech cipher
It sets.
Background technique
It with information technology continues to develop, quick payment has the characteristics that fast, easily, and main advantage is " fast ".
Quick payment has been popularized in middle big city.But safety of payment problem becomes hot spot concerned by people.Existing safety of payment is arranged
Apply fixed numerical ciphers, number and monogram password, dynamic password, the combination pin of numerical ciphers, number and letter
There will be the hidden danger of hacker's wooden horse theft user account password, the property or data of user there is a problem of stolen.And
Dynamic password is too time-consuming, and dynamic password has real-time, and dynamic password is not convenient enough, quick.
Summary of the invention
In view of this, it is an object of the invention to propose a kind of payment platform login method and dress based on speech cipher
It sets, solves the problems, such as that existing payment cipher safety is poor, not convenient and quick enough.
Based on the above-mentioned purpose payment platform login method provided by the invention based on speech cipher, comprising steps of
The audio signal is decomposed, establishes feature to each frame by the account and audio signal for receiving user's input
Point is to model;
It is inquired in preset user message table, obtains the characteristic point for including one or more foundation to model
Audio signal;
By the characteristic point of each audio signal found to model and newly-established all characteristic points to model progress
Match, obtains the corresponding account information of audio signal of successful match, judgement pair in the user message table if successful match
Whether the account information answered is identical as the account information of user's input, then logs in payment platform if they are the same, logs in if not identical
Failure;The direct login failure if matching is unsuccessful.
In some embodiments of the invention, after the audio signal for receiving user's input, further includes:
The audio signal received is sampled, the audio analog signals after sampling are converted into audio digital signals;
According to pre-set time threshold, to the audio digital signals framing windowing process;
Audio digital signals after each framing windowing process are subjected to time frequency processing;
The characteristic point of time-frequency spectrum after extraction scaling down processing, and characteristic point is established to model.
In some embodiments of the invention, the characteristic point of establishing is to model, comprising:
Using the peak curve of the first frame of audio signal as initial threshold curve, the threshold value of each frame after first frame
Curve can be multiplied by the threshold curve of former frame with the peak curve of the frame to be obtained multiplied by attenuation coefficient;
The peak point after screening is extracted according to the threshold curve of every frame;Wherein, by by every frame all peak points with
The frame threshold curve corresponding points are compared, and give up the peak point if peak point is lower than the corresponding points of threshold curve, if peak value
The corresponding points that point is higher than threshold curve then retain the peak point;
Choose the peak point before every frame after several screenings, the characteristic point as every frame;
Characteristic point in each characteristic point of former frame and its a later frame region is successively organized pair.
In some embodiments of the invention, the characteristic point of each audio signal that will be found to model and creates
Vertical all characteristic points match model, comprising:
According to the content information of characteristic point pair, all characteristic points of each audio signal own model with newly-established
Feature point model is corresponded;
Corresponding characteristic point is calculated to the time difference between model, the time difference is arranged according to sequence from small to large
It is poor to obtain minimum time for column;
Count the number that each audio signal has minimum time difference;
Judge whether the number is greater than or equal to preset minimum frequency threshold value, if more than or equal to preset minimum number
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success to threshold value;Or directly extraction has
The most audio signal of minimum time difference number, as with newly-built characteristic point to the successful audio signal of Model Matching.
In some embodiments of the invention, before the account and audio signal for receiving user's input, further includes:
Floating layer is popped up, the information of display reminding simultaneously starts monitoring process;
When monitoring the audio signal with input, the duration of starting monitoring process is obtained;
Judge whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure.
On the other hand, the present invention also provides a kind of payment platform entering device based on speech cipher, comprising:
Log-on message receiving unit, for receiving the account and audio signal of user's input.
Audio signal processing unit establishes characteristic point to model to each frame for decomposing the audio signal;
Query unit, for inquiring in preset user message table, acquisition includes one or more foundation
Audio signal of the characteristic point to model;
Matching unit, for by the characteristic point of each audio signal found to model and newly-established all characteristic points
Model is matched, obtains the corresponding account of audio signal of successful match in the user message table if successful match
Information judges whether corresponding account information is identical as the account information of user's input, payment platform is then logged in if they are the same, if not
Identical then login failure;The direct login failure if matching is unsuccessful.
In some embodiments of the invention, the audio signal processing unit, is also used to:
The audio signal received is sampled, the audio analog signals after sampling are converted into audio digital signals;
According to pre-set time threshold, to the audio digital signals framing windowing process;
Audio digital signals after each framing windowing process are subjected to time frequency processing;
The characteristic point of time-frequency spectrum after extraction scaling down processing, and characteristic point is established to model.
In some embodiments of the invention, the audio signal processing unit establishes characteristic point to model, comprising:
Using the peak curve of the first frame of audio signal as initial threshold curve, the threshold value of each frame after first frame
Curve can be multiplied by the threshold curve of former frame with the peak curve of the frame to be obtained multiplied by attenuation coefficient;
The peak point after screening is extracted according to the threshold curve of every frame;Wherein, by by every frame all peak points with
The frame threshold curve corresponding points are compared, and give up the peak point if peak point is lower than the corresponding points of threshold curve, if peak value
The corresponding points that point is higher than threshold curve then retain the peak point;
Choose the peak point before every frame after several screenings, the characteristic point as every frame;
Characteristic point in each characteristic point of former frame and its a later frame region is successively organized pair.
In some embodiments of the invention, the matching unit, by the characteristic point pair of each audio signal found
Model matches model with newly-established all characteristic points, comprising:
According to the content information of characteristic point pair, all characteristic points of each audio signal own model with newly-established
Feature point model is corresponded;
Corresponding characteristic point is calculated to the time difference between model, the time difference is arranged according to sequence from small to large
It is poor to obtain minimum time for column;
Count the number that each audio signal has minimum time difference;
Judge whether the number is greater than or equal to preset minimum frequency threshold value, if more than or equal to preset minimum number
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success to threshold value;Or directly extraction has
The most audio signal of minimum time difference number, as with newly-built characteristic point to the successful audio signal of Model Matching.
In some embodiments of the invention, the log-on message receiving unit receives the account and audio of user's input
Before signal, it is also used to:
Floating layer is popped up, the information of display reminding simultaneously starts monitoring process;
When monitoring the audio signal with input, the duration of starting monitoring process is obtained;
Judge whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure.
From the above it can be seen that the payment platform login method and device provided by the invention based on speech cipher,
By decomposing received audio signal, characteristic point is established to model to each frame;It inquires, obtains in preset user message table
Audio signal of the characteristic point to model of foundation must be included one or more;By the spy of each audio signal found
Sign point matches model with newly-established all characteristic points to model, obtains in the user message table if successful match
The corresponding account information of audio signal for obtaining successful match judges the account information whether corresponding account information inputs with user
It is identical, payment platform is then logged in if they are the same, the login failure if not identical;The direct login failure if matching is unsuccessful.To,
Payment platform login method and device of the present invention based on speech cipher can overcome the defect of existing payment cipher, make
Payment behavior is faster, more convenient, safer.
Detailed description of the invention
Fig. 1 is the flow diagram of the payment platform login method based on speech cipher in first embodiment of the invention;
Fig. 2 is the flow diagram for the payment platform login method that the present invention can refer in embodiment based on speech cipher;
Fig. 3 is the structural schematic diagram of the payment platform entering device based on speech cipher in the embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, and reference
Attached drawing, the present invention is described in more detail.
As shown in fig.1, for the process of the payment platform login method based on speech cipher in first embodiment of the invention
Schematic diagram, the payment platform login method based on speech cipher include:
Step 101, the account and audio signal of user's input are received.
Step 102, the audio signal is decomposed, characteristic point is established to model to each frame.
Step 103, it is inquired in preset user message table, obtains the characteristic point for including one or more foundation
To the audio signal of model.
Step 104, by the characteristic point of each audio signal found to model and newly-established all characteristic points to mould
Type is matched, if successful match thens follow the steps 105, if matching is unsuccessful to then follow the steps 106.
Step 105, the corresponding account information of audio signal of successful match, judgement pair are obtained in the user message table
Whether the account information answered is identical as the account information of user's input, then logs in payment platform if they are the same, executes if not identical
Step 106.
Step 106, login failure.
It, can be right after step 101 receives the account and audio signal of user's input in one preferably embodiment
The audio signal received is sampled.Preferably, 8kHz sample rate is taken to sample the audio-frequency information received.It will adopt
Audio analog signals after sample are converted into audio digital signals, then according to pre-set time threshold, to the audio
Digital signal framing windowing process.Preferably, pre-set time threshold is 10s, i.e., divides an audio number at interval of 10s
Word signal segment.Then, the audio digital signals after each framing windowing process are subjected to time frequency processing.Preferably, using quick
Fourier transformation (FFT) method is completed, the time of the audio digital signals after obtaining each framing windowing process and frequecy characteristic.
Finally, extracting the characteristic point of time-frequency spectrum after scaling down processing, and characteristic point is established to model.
The embodiment that can refer to as one, as shown in fig.2, the payment platform login method based on speech cipher
Following steps specifically can be used:
Step 201, the account and audio signal of user's input are received.
Wherein, user is according to the information input audio-frequency information of prompt, such as: any first song of your which favorite star
It is bent;The when and where that you meet for the first time with your lover;The date of birth etc. of your parent.Preferably, Mike can be passed through
Wind receives the audio-frequency information of user's input.
Step 202, the audio signal received is sampled.
The frequency range of common people's voice signal is 300Hz~3.4kHz, according to nyquist sampling theorem: ought only be adopted
When sample frequency is higher than twice of voice signal highest frequency, the voice signal that discrete analog signal indicates could be uniquely restored
At original sound.Preferably, 8kHz sample rate is taken to sample the audio-frequency information received in this embodiment.
Step 203, the audio analog signals after sampling are converted into audio digital signals.
As one embodiment, the audio analog signals after sampling can be converted by digital audio by analog-digital converter
Signal.
Step 204, according to pre-set time threshold, to the audio digital signals framing windowing process.
In one embodiment it would be required that a length of 10s when voice, that is to say, that pre-set time threshold is 10s, i.e.,
An audio digital signals section is divided at interval of 10s.
Step 205, the audio digital signals after each framing windowing process are subjected to time frequency processing.Specific implementation includes:
Using Fast Fourier Transform (FFT) (FFT) method complete, the time of the audio digital signals after obtaining each framing windowing process and
Frequecy characteristic.
Step 206, the characteristic point for extracting time-frequency spectrum after scaling down processing, and establishes characteristic point to model.Specific implementation process
Include:
Step 1: using the peak curve of the first frame of audio signal as initial threshold curve.
Step 2: the threshold curve of each frame can pass through the threshold curve of former frame and the peak value of the frame after first frame
Curve multiplication is obtained multiplied by attenuation coefficient, and wherein attenuation coefficient takes 0.98.
Step 3: the peak point after screening is extracted according to the threshold curve of every frame.Wherein, by by all peaks in every frame
Value point is compared with the frame threshold curve corresponding points, gives up the peak point if peak point is lower than the corresponding points of threshold curve,
Retain the peak point if the corresponding points that peak point is higher than threshold curve.
Step 4: the peak point before every frame after several screenings, the characteristic point of as every frame are chosen.In embodiment, it selects
Take the peak point before every frame after 5 screenings.
Step 5: the characteristic point in each characteristic point of former frame and its a later frame region is successively organized pair, every a pair
Characteristic point all uses one 32 lint-long integers according to description, 32 data by two characteristic points frequency and two characteristic points
Time difference composition.
Step 207, it is inquired in user message table, obtains the characteristic point for including one or more foundation to model
Audio signal.
Step 208, by the characteristic point of each audio signal found to model and newly-established all characteristic points to mould
Type is matched, if successful match thens follow the steps 209, if matching is unsuccessful to then follow the steps 210.
Preferably, can using hash reverse index method to all characteristic points found to model with it is newly-established
Characteristic point matches model.Further, specific implementation process includes:
Step 1: according to the content information of characteristic point pair, all characteristic points of each audio signal to model and are created
Vertical all feature point models are corresponded.Such as: content of all characteristic points of an audio signal to model are as follows:
" I ", "Yes", " small ", " king ", " son ", and newly-established all characteristic points are to the content of model: " I ", "Yes", " small ",
" public affairs ", " master ", then " I " of audio signal, "Yes", " small " characteristic point correspond upper newly-established characteristic point pair to model
" I " of model, "Yes", " small ".
Step 2: corresponding characteristic point is calculated to the time difference between model, by the time difference according to sequence from small to large
It is arranged, it is poor to obtain minimum time.
Step 3: the number that each audio signal has minimum time difference is counted.
Step 4: judging whether the number is greater than or equal to preset minimum frequency threshold value, if more than or be equal to it is preset
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success to minimum frequency threshold value.Or directly
The audio signal for having minimum time difference number most is extracted, the successful audio of Model Matching is believed as with newly-built characteristic point
Number.
Step 209, the corresponding account information of audio signal of successful match, judgement pair are obtained in the user message table
Whether the account information answered is identical as the account information of user's input, then logs in payment platform if they are the same, executes if not identical
Step 210.
Step 210, login failure.
It, can be with it should also be noted that, when user did not register (i.e. in user message table there is no the user)
By step 201 to step 206, then the account of the user stores to pre-set model with corresponding characteristic point
In user message table.That is, the process that user logs in is step 201 to step 210.And the process of user's registration is step
201, to step 206, then store the account of the user to model to pre-set user information with corresponding characteristic point
In table.
As example is further carried out, when step 209 does not have in pre-set user message table according to the account
When finding identical as the account information of user's input, floating layer can be popped up, new account is prompted to register.New account is registered when receiving
Number instruction when, the account and the characteristic point can be established in preset user message table to the mapping relations of model.It is excellent
Selection of land can directly store the mapping relations of the new account Yu the feature point model when registering new account.It can also first set
Prompt information when login is set, then acquires the audio-frequency information of the prompt information, and this is obtained according to step 202 to step 206
The characteristic point of the audio-frequency information of secondary acquisition to model, by the new account and the characteristic point that this acquisition obtains to model together with store
Into the user message table.
In one preferably embodiment, when prompt information when logging in is arranged, answer prompt information can also be set
Duration threshold value.When executing the duration threshold value of setting, specifically implementation process includes:
Step 1: pop-up floating layer, the information of display reminding simultaneously start monitoring process.Wherein, the prompt information is to need
The prompt information for the audio signal to be inputted, such as: prompt information are as follows: " what a nearest favorite book is? "
Step 2: when monitoring the audio signal with input, the duration of starting monitoring process is obtained.
Step 3: judging whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure, if small
In or equal to thening follow the steps 202 to 210.
Preferably, the number for answering prompt information can also be set, such as three times, it can provide chance three times and carry out sound
The input of frequency information.In addition, not all being successfully logged onto payment platform then after answering the number of prompt information by setting
The account can be lockked, with the Anti-theft account.At the same time it can also design the unlock setting to the account, such as progress body
Part verifying (can be and upload identity card).Preferably, it can also be stored with account in pre-set user message table and use
The mapping relations of family information, can be by comparing user information when being unlocked to account.
In another aspect of this invention, a kind of payment platform entering device based on speech cipher is additionally provided, such as Fig. 3 institute
Show, the payment platform entering device based on speech cipher includes sequentially connected log-on message receiving unit 301, audio letter
Number processing unit 302, query unit 303, matching unit 304.Wherein, log-on message receiving unit 301 receives user's input
Account and audio signal, audio signal processing unit 302 decomposes the audio signal, establishes characteristic point pair to each frame
Model.Query unit 303 is inquired in preset user message table, obtains the characteristic point for including one or more foundation
To the audio signal of model.Matching unit 304 is by the characteristic point of each audio signal found to model and newly-established institute
There is characteristic point to match model, obtains the audio signal pair of successful match in the user message table if successful match
The account information answered judges whether corresponding account information is identical as the account information of user's input, then logs in payment if they are the same
Platform, the login failure if not identical;The direct login failure if matching is unsuccessful.
In one preferably embodiment, the audio signal processing unit 302 can to the audio signal received into
The following processing of row: the audio signal received can be sampled.Preferably, take 8kHz sample rate to the audio received
Information is sampled.Audio analog signals after sampling are converted into audio digital signals, then according to the pre-set time
Threshold value, to the audio digital signals framing windowing process.Preferably, pre-set time threshold be 10s, i.e., at interval of
10s divides an audio digital signals section.Then, the audio digital signals after each framing windowing process are carried out at time-frequency
Reason.Preferably, it is completed using Fast Fourier Transform (FFT) (FFT) method, the digital audio letter after obtaining each framing windowing process
Number time and frequecy characteristic.Finally, extracting the characteristic point of time-frequency spectrum after scaling down processing, and characteristic point is established to model.
Further embodiment, when audio signal processing unit 302 establishes characteristic point to model, specific implementation process
Include:
Step 1: using the peak curve of the first frame of audio signal as initial threshold curve.
Step 2: the threshold curve of each frame can pass through the threshold curve of former frame and the peak value of the frame after first frame
Curve multiplication is obtained multiplied by attenuation coefficient, and wherein attenuation coefficient takes 0.98.
Step 3: the peak point after screening is extracted according to the threshold curve of every frame.Wherein, by by all peaks in every frame
Value point is compared with the frame threshold curve corresponding points, gives up the peak point if peak point is lower than the corresponding points of threshold curve,
Retain the peak point if the corresponding points that peak point is higher than threshold curve.
Step 4: the peak point before every frame after several screenings, the characteristic point of as every frame are chosen.In embodiment, it selects
Take the peak point before every frame after 5 screenings.
Step 5: the characteristic point in each characteristic point of former frame and its a later frame region is successively organized pair, every a pair
Characteristic point all uses one 32 lint-long integers according to description, 32 data by two characteristic points frequency and two characteristic points
Time difference composition.
As in another embodiment of the present apparatus, the matching unit 304 is by the spy of each audio signal found
When sign point matches model with newly-established all characteristic points to model, specific implementation process includes:
Step 1: according to the content information of characteristic point pair, all characteristic points of each audio signal to model and are created
Vertical all feature point models are corresponded.
Step 2: corresponding characteristic point is calculated to the time difference between model, by the time difference according to sequence from small to large
It is arranged, it is poor to obtain minimum time.
Step 3: the number that each audio signal has minimum time difference is counted.
Step 4: judging whether the number is greater than or equal to preset minimum frequency threshold value, if more than or be equal to it is preset
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success to minimum frequency threshold value.Or directly
The audio signal for having minimum time difference number most is extracted, the successful audio of Model Matching is believed as with newly-built characteristic point
Number.
In addition, the embodiment that can refer to as one, prompt when logging in is being arranged in the log-on message receiving unit 301
When information, the duration threshold value for answering prompt information can also be set.When executing the duration threshold value of setting, specifically implementation process
Include:
Step 1: pop-up floating layer, the information of display reminding simultaneously start monitoring process.Wherein, the prompt information is to need
The prompt information for the audio signal to be inputted, such as: prompt information are as follows: " what a nearest favorite book is? "
Step 2: when monitoring the audio signal with input, the duration of starting monitoring process is obtained.
Step 3: judging whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure.
It should be noted that in the specific implementation of the payment platform entering device of the present invention based on speech cipher
Hold, has been described in detail in the payment platform login method described above based on speech cipher, therefore in this duplicate contents
No longer illustrate.
In conclusion payment platform login method, the device provided by the invention based on speech cipher, creatively;Together
When, it is mainly used in the safeguard protection of quick payment.Speech cipher can effectively take precautions against hacker's wooden horse theft user account password,
The multiple networks problem such as false website leads to the property of user or the loss of data.Speech cipher ensures the peace of user's payment
Quan Xing;Also, convenient, quick, high safety that the present invention has many advantages, such as;Moreover, having extensive, great dissemination;Finally,
Entire described payment platform login method and device based on speech cipher are compact, easily controllable.
It should be understood by those ordinary skilled in the art that: the discussion of any of the above embodiment is exemplary only, not
It is intended to imply that the scope of the present disclosure (including claim) is limited to these examples;Under thinking of the invention, above embodiments
Or can also be combined between the technical characteristic in different embodiments, step can be realized with random order, and be existed such as
Many other variations of the upper different aspect of the invention, for simplicity, they are not provided in details.
In addition, to simplify explanation and discussing, and in order not to obscure the invention, it can in provided attached drawing
It is connect with showing or can not show with the well known power ground of integrated circuit (IC) chip and other components.Furthermore, it is possible to
Device is shown in block diagram form, to avoid obscuring the invention, and this has also contemplated following facts, i.e., about this
The details of the embodiment of a little block diagram arrangements be height depend on will implementing platform of the invention (that is, these details should
It is completely within the scope of the understanding of those skilled in the art).Elaborating that detail (for example, circuit) is of the invention to describe
In the case where exemplary embodiment, it will be apparent to those skilled in the art that can be in these no details
In the case where or implement the present invention in the case that these details change.Therefore, these descriptions should be considered as explanation
Property rather than it is restrictive.
Although having been incorporated with specific embodiments of the present invention, invention has been described, according to retouching for front
It states, many replacements of these embodiments, modifications and variations will be apparent for those of ordinary skills.Example
Such as, discussed embodiment can be used in other memory architectures (for example, dynamic ram (DRAM)).
The embodiment of the present invention be intended to cover fall into all such replacements within the broad range of appended claims,
Modifications and variations.Therefore, all within the spirits and principles of the present invention, any omission, modification, equivalent replacement, the improvement made
Deng should all be included in the protection scope of the present invention.
Claims (6)
1. a kind of payment platform login method based on speech cipher, which is characterized in that comprising steps of
Receive the account and audio signal of user's input;
The audio signal received is sampled, the audio analog signals after sampling are converted into audio digital signals;According to
Pre-set time threshold, to the audio digital signals framing windowing process;By the sound after each framing windowing process
Frequency digital signal carries out time frequency processing;The characteristic point of time-frequency spectrum after extraction scaling down processing, and characteristic point is established to model;Wherein,
The characteristic point of establishing is to model, comprising: using the peak curve of the first frame of audio signal as initial threshold curve, first
The threshold curve of each frame can be multiplied with the peak curve of the frame multiplied by decaying by the threshold curve of former frame after frame
Coefficient obtains;The peak point after screening is extracted according to the threshold curve of every frame;Wherein, by by every frame all peak points with
The frame threshold curve corresponding points are compared, and give up the peak point if peak point is lower than the corresponding points of threshold curve, if peak value
The corresponding points that point is higher than threshold curve then retain the peak point;The peak point before every frame after several screenings is chosen, as every frame
Characteristic point;Characteristic point in each characteristic point of former frame and its a later frame region is successively organized pair;
It is inquired in preset user message table, obtains audio of the characteristic point to model for including one or more foundation
Signal;
The characteristic point of each audio signal found matches model with newly-established all characteristic points to model, if
Successful match then obtains the corresponding account information of audio signal of successful match in the user message table, judges corresponding account
Whether number information is identical as the account information of user's input, then logs in payment platform if they are the same, the login failure if not identical;If
Match unsuccessful then direct login failure.
2. the method according to claim 1, wherein the characteristic point pair of each audio signal that will be found
Model matches model with newly-established all characteristic points, comprising:
According to the content information of characteristic point pair, by all characteristic points of each audio signal to model and newly-established all features
Point model is corresponded;
Corresponding characteristic point is calculated to the time difference between model, the time difference is arranged according to sequence from small to large, is obtained
It is poor to obtain minimum time;
Count the number that each audio signal has minimum time difference;
Judge whether the number is greater than or equal to preset minimum frequency threshold value, if more than or equal to preset minimum frequency threshold value
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success;Or it directly extracts with minimum
The most audio signal of time difference number, as with newly-built characteristic point to the successful audio signal of Model Matching.
3. method according to claim 1 or 2, which is characterized in that the account for receiving user's input and audio letter
Before number, further includes:
Floating layer is popped up, the information of display reminding simultaneously starts monitoring process;
When monitoring the audio signal with input, the duration of starting monitoring process is obtained;
Judge whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure.
4. a kind of payment platform entering device based on speech cipher characterized by comprising
Log-on message receiving unit, for receiving the account and audio signal of user's input;
Audio signal processing unit turns the audio analog signals after sampling for sampling to the audio signal received
Change audio digital signals into;According to pre-set time threshold, to the audio digital signals framing windowing process;It will be every
Audio digital signals after a framing windowing process carry out time frequency processing;The characteristic point of time-frequency spectrum after extraction scaling down processing, and build
Characteristic point is found to model;Wherein, the characteristic point of establishing is to model, comprising: makees the peak curve of the first frame of audio signal
For initial threshold curve, the threshold curve of each frame can pass through the threshold curve of former frame and the peak of the frame after first frame
The multiplication of value curve is obtained multiplied by attenuation coefficient;The peak point after screening is extracted according to the threshold curve of every frame;Wherein, pass through by
All peak points in every frame are compared with the frame threshold curve corresponding points, if peak point is lower than the corresponding points of threshold curve
Give up the peak point, retains the peak point if the corresponding points that peak point is higher than threshold curve;Several are screened before choosing every frame
Peak point afterwards, the characteristic point as every frame;By the characteristic point in each characteristic point of former frame and its a later frame region according to
Secondary group pair;
Query unit obtains the feature for including one or more foundation for inquiring in preset user message table
Audio signal of the point to model;
Matching unit, for by the characteristic point of each audio signal found to model and newly-established all characteristic points to mould
Type is matched, and obtains the corresponding account letter of audio signal of successful match in the user message table if successful match
Breath judges whether corresponding account information is identical as the account information of user's input, payment platform is then logged in if they are the same, if not phase
Same then login failure;The direct login failure if matching is unsuccessful.
5. device according to claim 4, which is characterized in that the matching unit, each audio signal that will be found
Characteristic point model matches model with newly-established all characteristic points, comprising:
According to the content information of characteristic point pair, by all characteristic points of each audio signal to model and newly-established all features
Point model is corresponded;
Corresponding characteristic point is calculated to the time difference between model, the time difference is arranged according to sequence from small to large, is obtained
It is poor to obtain minimum time;
Count the number that each audio signal has minimum time difference;
Judge whether the number is greater than or equal to preset minimum frequency threshold value, if more than or equal to preset minimum frequency threshold value
Then otherwise the audio signal and newly-built characteristic point match unsuccessful Model Matching success;Or it directly extracts with minimum
The most audio signal of time difference number, as with newly-built characteristic point to the successful audio signal of Model Matching.
6. device according to claim 4 or 5, which is characterized in that it is defeated that the log-on message receiving unit receives user
Before the account and audio signal that enter, it is also used to:
Floating layer is popped up, the information of display reminding simultaneously starts monitoring process;
When monitoring the audio signal with input, the duration of starting monitoring process is obtained;
Judge whether the duration obtained is greater than preset duration threshold value, if more than the then secondary login failure.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610703600.3A CN106384595B (en) | 2016-08-22 | 2016-08-22 | A kind of payment platform login method and device based on speech cipher |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610703600.3A CN106384595B (en) | 2016-08-22 | 2016-08-22 | A kind of payment platform login method and device based on speech cipher |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106384595A CN106384595A (en) | 2017-02-08 |
CN106384595B true CN106384595B (en) | 2019-04-02 |
Family
ID=57916850
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610703600.3A Active CN106384595B (en) | 2016-08-22 | 2016-08-22 | A kind of payment platform login method and device based on speech cipher |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106384595B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108768977A (en) * | 2018-05-17 | 2018-11-06 | 东莞市华睿电子科技有限公司 | A kind of terminal system login method based on speech verification |
CN108876983A (en) * | 2018-05-17 | 2018-11-23 | 东莞市华睿电子科技有限公司 | A kind of unlocking method of safety box with function of intelligent lock |
CN111883141B (en) * | 2020-07-27 | 2022-02-25 | 重庆金宝保信息技术服务有限公司 | Text semi-correlation voiceprint recognition method and system |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101772015A (en) * | 2008-12-29 | 2010-07-07 | 卢中江 | Method for starting up mobile terminal through voice password |
WO2011146531A1 (en) * | 2010-05-18 | 2011-11-24 | Acea Biosciences, Inc | Data analysis of impedance-based cardiomyocyte-beating signals as detected on real-time cell analysis (rtca) cardio instruments |
CN102098159A (en) * | 2010-07-28 | 2011-06-15 | 胡旭光 | Secret key device and method for mobile phone |
CN103685185B (en) * | 2012-09-14 | 2018-04-27 | 上海果壳电子有限公司 | Mobile equipment voiceprint registration, the method and system of certification |
CN104022879B (en) * | 2014-05-29 | 2018-06-26 | 金蝶软件(中国)有限公司 | The method and device of voice safety check |
-
2016
- 2016-08-22 CN CN201610703600.3A patent/CN106384595B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN106384595A (en) | 2017-02-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108039176B (en) | Voiceprint authentication method and device for preventing recording attack and access control system | |
CN102254559A (en) | Identity authentication system and method based on vocal print | |
US8793135B2 (en) | System and method for auditory captchas | |
CN106961418A (en) | Identity identifying method and identity authorization system | |
CN105306657B (en) | Personal identification method, device and communicating terminal | |
CN106384595B (en) | A kind of payment platform login method and device based on speech cipher | |
US8812319B2 (en) | Dynamic pass phrase security system (DPSS) | |
CN103685185B (en) | Mobile equipment voiceprint registration, the method and system of certification | |
CN107886958A (en) | Express cabinet pickup method and device based on voiceprint | |
CN108922518A (en) | voice data amplification method and system | |
CN106782572A (en) | The authentication method and system of speech cipher | |
CN105913850B (en) | Text correlation vocal print method of password authentication | |
CN108074310A (en) | Voice interactive method and intelligent lock administration system based on sound identification module | |
CN107886957A (en) | Voice wake-up method and device combined with voiceprint recognition | |
CN110459204A (en) | Audio recognition method, device, storage medium and electronic equipment | |
CN109120605A (en) | Authentication and account information variation and device | |
AU2017305245A1 (en) | Call classification through analysis of DTMF events | |
CN110178179A (en) | Voice signature for being authenticated to electronic device user | |
CN110061984A (en) | Account switching method, onboard system and the vehicle of onboard system | |
CN104468522A (en) | Voiceprint authentication method and device | |
WO2005022396A1 (en) | Mutual authentication system between user and system | |
CN108062464A (en) | Terminal control method and system based on Application on Voiceprint Recognition | |
WO2018129869A1 (en) | Voiceprint verification method and apparatus | |
CN105047192B (en) | Statistics phoneme synthesizing method based on Hidden Markov Model and device | |
CN105700897B (en) | A kind of method, apparatus and terminal device starting application program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |