WO2017201972A1 - 基于脑电信号的身份识别的方法和装置 - Google Patents

基于脑电信号的身份识别的方法和装置 Download PDF

Info

Publication number
WO2017201972A1
WO2017201972A1 PCT/CN2016/104445 CN2016104445W WO2017201972A1 WO 2017201972 A1 WO2017201972 A1 WO 2017201972A1 CN 2016104445 W CN2016104445 W CN 2016104445W WO 2017201972 A1 WO2017201972 A1 WO 2017201972A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
segment
ssvep
ssvep signal
correlation coefficient
Prior art date
Application number
PCT/CN2016/104445
Other languages
English (en)
French (fr)
Inventor
袁鹏
薛希俊
姚骏
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2017201972A1 publication Critical patent/WO2017201972A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing

Definitions

  • the present invention relates to the field of information technology, a method and apparatus for identity recognition based on EEG signals.
  • Biometrics technology generally refers to the use of some physiological characteristics or behavioral characteristics inherent in the human body for identification.
  • the physiological characteristics of the human body generally include: human face, fingerprint, palm shape, iris, etc.; human behavior characteristics may include: notes, gait, and the like.
  • the existing biometric technology also faces some problems. For example, for identification by fingerprint, a fake finger made of gelatin can successfully fool the fingerprint recognition system; for identification by iris, the false iris feature etched on the contact lens can make the iris recognition system indistinguishable. true and false. These problems pose new challenges to biometrics and inspire people to explore new biometrics.
  • EEG electronic medical record
  • studies have shown that even under the same external stimulus or when people are thinking about the same problem, the brain electrical signals induced by the brains of different subjects are different, that is, EEG has significant individual differences.
  • EEG has many advantages such as difficulty in copying and forgery, and can be modulated by the subject's own attention. Therefore, a series of identification methods based on various modes of EEG have emerged. For example, an identification method based on resting EEG, an identification method based on an imaginary motion state EEG feature, an EEG identification method based on a P300 event-related potential, and the like.
  • the signal-to-noise ratio of the EEG signals used above is generally low, and the signal characteristics are not stable enough. It is usually necessary to collect multiple lead EEG signals and require a large number of training samples, which is not convenient to use. For example, based on the identification method of resting EEG, the spontaneous EEG in resting state has a high degree of non-stationaryness, and is also susceptible to the state of the individual, and the variability within the individual is large; and in testing, the user generally needs The EEG signal is collected by a plurality of electrodes covering the whole brain, which is inconvenient for the user.
  • the signal-to-noise ratio of the P300 event-related potential is very low, and a large number of repeated stimuli are required and the average is superimposed to obtain a stable waveform, which has a long authentication time and is inconvenient to use;
  • the P300 event-related potential involves the high-level cognitive function of human beings, which is mostly induced by novel stimuli and is susceptible to the mental state of the user's subject. Long-term stimulation will cause the user to have an adaptive response, resulting in induced P300 event-related potential attenuation. , It is not conducive to the extraction of signal features and has poor reliability.
  • the present application provides a method and apparatus for identity recognition based on an EEG signal, which can perform identity recognition according to a steady-state visual evoked potential SSVEP signal, thereby improving the reliability of identity recognition.
  • a method for identity recognition based on an EEG signal comprising: determining a target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], n being a positive integer;
  • the user to be detected displays the n-segment stimulation signal corresponding to the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . .
  • the i-th frequency f i , i in the frequency sequence [f 1 , f 2 , f 3 , ..., f n ] takes 1 , 2 , 3 , ...
  • n acquires the user to be detected due to the n-segment stimulation signal
  • the n-stage steady-state visual evoked potential SSVEP signal when the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal is greater than or equal to the threshold, determining that the identity of the user to be detected is correct; when the n-segment SSVEP signal and the n-segment When the similarity of the preset SSVEP signal is less than the threshold, it is determined that the identity of the user to be detected is incorrect.
  • the method for identifying the EEG signal based on the present application displays a stimulus signal for the user to be detected according to the target stimulation frequency sequence, thereby collecting the SSVEP signal generated by the user to be detected, and comparing the SSVEP signal with the preset SSVEP signal.
  • the identification of the user to be detected is performed. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • the SSVEP signal is used for identification, and the SSVEP signal is generated by a stable periodic visual stimulus, that is, whether the preset SSVEP signal previously recorded by the original user or the user to be detected is identified during the identification.
  • the acquired SSVEP signals require stable periodic visual stimuli to induce SSVEP signals in the human brain. Therefore, it is necessary to determine the frequency of at least one visual stimulus.
  • the user when the SSVEP signal is acquired, the user may be provided with a wearable dry electrode cap that includes at least one dry electrode placed at the scalp of the user's occipital region to facilitate acquisition of the user's SSVEP signal.
  • a wearable dry electrode cap that includes at least one dry electrode placed at the scalp of the user's occipital region to facilitate acquisition of the user's SSVEP signal.
  • the wearable EEG device ie, the wearable dry electrode cap.
  • the collected SSVEP signals can be wirelessly transmitted to the corresponding signal analysis area for subsequent analysis and processing.
  • the determining the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] includes: acquiring the to-be-detected user input to be detected Stimulus frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ]; when the stimulus frequency sequence to be detected [f' 1 , f' 2 , f' 3 , ..., f' n ] stimulation frequency to the target sequence [f 1, f 2, f 3, ; to acquire the target sequence of stimulation frequencies [f 1, f 2, f 3, ...., f' n] corresponding to The n-segment stimulation signal; when the stimulation frequency sequence to be detected [f' 1 , f' 2 , f' 3 , ..., f' n ] and the target stimulation frequency sequence [f 1 , f
  • the identification is performed, so that the exclusive stimulation sequence plus the SSVEP signal dual authentication mode improves the level of device confidentiality and is less likely to be invaded.
  • the range of each frequency in the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] set by the user is generally between 6 Hz and 100 Hz, and the n frequencies may be in accordance with the size. Arranged sequentially, or randomly, the n frequencies may include at least two identical frequencies.
  • each frequency in the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] may be arranged in order or randomly arranged, therefore, in determining the stimulation frequency sequence [f' 1 , f′ 2 , f′ 3 , . . . , f′ n ] may be the same as the target stimulation sequence [f 1 , f 2 , f 3 , . . . , f n ], and may include determining each of the two sequences correspondingly to each frequency The order is the same, the corresponding frequency of each frequency is the same, and the number of frequencies in the two sequences is the same.
  • the user input stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is When the target stimulation sequence is a subset of [f 1 , f 2 , f 3 , ..., f n ], the user input stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ... , f' n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ].
  • the user to be detected inputs the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ], and determines the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ... , if f′ n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , . . . , f n ], the input order may be disregarded, that is, the size of each frequency input corresponds to the target stimulation sequence.
  • the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] and the target stimulation sequence [f 1 , f are determined. 2 , f 3 , ..., f n ] are the same.
  • the user to be detected inputs the stimulation frequency sequence [f' 1 , f′ 2 , f′ 3 , . . . , f′ n ], and determines the stimulation frequency sequence [f′ 1 , f′ 2 , f′ 3 , ..., if f' n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ], the number of inputs, that is, the size and target of each frequency input, may be disregarded.
  • the stimulation frequency sequence can be determined when the corresponding frequency in the stimulation sequence is the same and the order of the input frequency is the same as the frequency sequence in the target stimulation sequence, or only if the magnitude of each frequency input is the same as the corresponding frequency in the target stimulation sequence.
  • f' 1 , f' 2 , f' 3 , ..., f' n ] are identical to the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ].
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] may also not be used as an identification basis.
  • the identification device automatically acquires the prior The user-set target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ], or randomly acquire or sequentially acquire a part of a series of frequencies set in advance by the user to form a target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], by the acquired target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], the user is detected to display the stimulation signal, and is collected.
  • the SSVEP signal generated by the user due to the stimulation signal is detected, and the identification of the SSVEP signal is performed.
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] is displayed for the user to be detected.
  • the n-segment stimulation signal includes: displaying n-segment stimulation signals for the user to be detected, wherein the first segment stimulation signal is displayed for a period of time at a frequency of f 1 , and the second segment stimulation signal is displayed for a period of time at a frequency of f 2 , the third segment The stimulation signal is displayed at a frequency of f 3 for a period of time until the nth segment of the stimulation signal is displayed at a frequency of f n for a period of time.
  • the n-segment SSVEP signal generated by the user to be detected according to the n-segment stimulation signal may be collected.
  • the n-segment SSVEP signal is generated according to the n-segment stimulation signals corresponding to the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ].
  • the n-segment SSVEP signal is collected and saved as a preset SSVEP signal.
  • the stimulation signal may be a picture with a certain pattern, the picture is displayed according to the corresponding frequency f i ; or the stimulation signal may also be a light, and the frequency of the switch or the alternating light and light of the light is the display frequency f i ;
  • the stimulation signal may also be any other form that can induce the SSVEP signal generated by the human brain, and is displayed to the user through the frequency f i in order to obtain the SSVEP signal generated by the response of the user to be detected to the stimulation signal.
  • the method further includes: the n-segment SSVEP signal according to the at least one domain and the n-segment preset SSVEP of the corresponding at least one domain a signal, determining a correlation coefficient vector between the n-segment SSVEP signal and the n-segment preset SSVEP signal Correlation coefficient vector
  • Each element in the representation represents a correlation coefficient between the n-segment SSVEP signal and the n-segment preset SSVEP signal
  • Each element in the representation represents the correlation coefficient vector The weight value of the corresponding element in .
  • the weight parameter vector Each element in the corresponding coefficient vector The weight of the correlation coefficient represented by each element.
  • the weight parameter vector Each element in the set is set equal to among them, Representation correlation coefficient vector The number of elements included in the. Or when the correlation coefficient vector When some elements are relatively small, it indicates that the element can strongly reflect the distinguishability between the real user and the non-real user, so the weight value corresponding to the element can be modulated to be too large.
  • the identity is determined according to the similarity between the n-segment SSVEP signal of the user to be detected and the n-segment preset SSVEP signal.
  • the similarity is greater than or equal to the threshold, it indicates that the identity of the user to be detected is correct, that is, the identification based on the SSVEP signal is used; when the similarity is less than the threshold, the identity of the user to be detected is incorrect, that is, cannot pass this time.
  • Identification based on SSVEP signals may be set according to actual conditions, and the threshold may be continuously updated according to multiple tests.
  • determining a correlation coefficient vector between the n-segment SSVEP signal and the n-segment preset SSVEP signal includes: determining a time domain correlation coefficient of the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal and X′ i (f i , T i ) in the n-segment preset SSVEP signal is r X (f i ),i takes 1, 2, 3...n; according to the time domain correlation coefficient r X (f i ), the correlation coefficient vector is determined
  • determining a correlation coefficient vector between the n-segment SSVEP signal and the n-segment preset SSVEP signal includes: transforming the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , Y i ); and determining X′ i in the n-segment SSVEP signal ( f i , T i ) transform to obtain Z′ i (f i , Y i ); determine that the transform domain correlation coefficient of Z i (f i , Y i ) and Z′ i (f i , Y i ) is r Y ( f i ),i takes 1, 2, 3...n; determines the correlation coefficient vector according to the transform domain correlation coefficient r Y (f i )
  • determining a correlation coefficient vector between the n-segment SSVEP signal and the n-segment preset SSVEP signal includes: Fourier transforming the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , F i ); and determining the X in the n-segment preset SSVEP signal ' i (f i , T i ) performs Fourier transform to obtain Z′ i (f i , F i ); determines the frequency domain of Z i (f i , F i ) and Z′ i (f i , F i )
  • the correlation coefficient is r F (f i ), i is 1, 2, 3...n; the correlation coefficient vector is determined according to the frequency domain correlation coefficient r F (f i ),
  • X i (f i , T i ) or X' i (f i , T i ) may be a matrix.
  • X i (f i , T i ) or X′ i (f i , T i ) can be reduced to one-dimensional by classical data dimension reduction methods such as principal component analysis (PCA) or canonical correlation analysis (CCA).
  • PCA principal component analysis
  • CCA canonical correlation analysis
  • Time domain signal The present application is described by taking X i (f i , T i ) and X′ i (f i , T i ) as one-dimensional time domain signals as an example.
  • transform domain correlation coefficient r y (f i ) may include a frequency domain correlation coefficient r F (f i ).
  • the time domain correlation coefficient is r X (f i ) and the transform domain correlation coefficient is r Y (f i ), and both may be linear correlation coefficients, for example, Pearson linear correlation coefficients.
  • determining a correlation coefficient vector according to n time domain correlation coefficients r X (f i ) and/or n transform domain correlation coefficients r Y (f i ) Correlation coefficient vector The elements in the medium can represent time domain correlation coefficients, and can also represent frequency domain correlation coefficients.
  • the correlation coefficient vector N elements may be included, and the n elements may include i time domain correlation coefficients r X (f i ), and (ni) transform domains including remaining (ni) time domain correlation coefficients r X (f i )
  • the correlation coefficient is r Y (f i ); for example, the 2n elements include n time domain correlation coefficients r X (f i ) and the n transform domain correlation coefficients are r Y (f i ).
  • an apparatus for identity recognition based on an EEG signal for performing the method of any of the above-described first aspect or any of the possible implementations of the first aspect.
  • the apparatus comprises means for performing the method of any of the above-described first aspect or any of the possible implementations of the first aspect.
  • an apparatus for identification based on an EEG signal comprising: a storage unit and a processor, the storage unit is configured to store an instruction, the processor is configured to execute the instruction stored by the memory, and when the processor When the instructions stored by the memory are executed, the execution causes the processor to perform the method of the first aspect or any of the possible implementations of the first aspect.
  • a computer readable medium for storing a computer program comprising instructions for performing the method of the first aspect or any of the possible implementations of the first aspect.
  • FIG. 1 is a schematic flow chart of a method for identification based on EEG signals according to an embodiment of the present invention.
  • FIG. 2 is a schematic block diagram of an apparatus for identification based on EEG signals in accordance with an embodiment of the present invention.
  • FIG. 3 is a schematic block diagram of an apparatus for identity recognition based on an EEG signal according to another embodiment of the present invention.
  • FIG. 1 shows a schematic flow diagram of a method 100 for EEG-based identity recognition in accordance with an embodiment of the present invention.
  • the method 100 can be performed by a device that requires authentication, for example, an identity device installed on a safe can be authenticated in accordance with the method 100.
  • the method 100 includes:
  • the method for identifying an EEG based signal needs to collect and save a user's EEG signal as a preset EEG signal in advance, and collect the brain of the user to be detected when performing identity verification or identity recognition.
  • the electrical signal compares the EEG signal with the saved preset EEG signal. If it matches the preset EEG signal, it indicates that the identity of the user to be detected is correct, and the identity verification passes; if it does not match the preset EEG, it indicates that Detecting user identity errors, authentication failed.
  • the EEG signal obtained by the identity verification in the embodiment of the present invention refers to a Steady-State Visual Evoked Potentials (SSVEP) signal, and the identity is identified according to the SSVEP signal.
  • SSVEP signal is generally induced by a stable periodic visual stimulus, which is typically recorded in the occipital region of the human brain.
  • the SSVEP signal mainly contains the EEG component of the same frequency as the stimulation signal, and the harmonic components and other frequency components are also included in the signal due to the nonlinear characteristics of the visual system and the influence of the spontaneous EEG.
  • EEG Event Related Potentials
  • VEP Visual Evoked Potentials
  • the SSVEP signal is generated by a stable periodic visual stimulus, that is, whether the preset SSVEP signal input by the user or the SSVEP signal to be detected by the user needs a stable cycle.
  • sexual visual stimulation induces the generation of SSVEP signals in the human brain. Therefore, it is necessary to determine the frequency of at least one visual stimulus.
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] needs to be selected first, and n is a positive integer, and each of the target stimulation frequency sequences
  • the frequency corresponds to the display frequency of the stimulation signal.
  • the frequency f i in the target stimulation frequency sequence refers to the display frequency of the i-th stimulation signal, that is, f i , thereby obtaining the user's preset SSVEP signal for the i-th stimulation signal.
  • the user here refers to the original user at the time of identification, for example, when the safe is opened, the holder of the safe can be the user in the embodiment of the present invention, and the user sets the preset SSVEP.
  • the signal is used as the basis for identification.
  • the range of each frequency in the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] set by the user is generally between 6 Hz and 100 Hz, and the n frequencies may be in accordance with the size.
  • the user can set the target stimulation frequency sequence to [20, 57, 38, 64], and the embodiment of the present invention is not limited thereto.
  • at least two identical frequency values may be included in the n frequencies, that is, the n frequency values may be repeated.
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ] is first determined. Since the target stimulation frequency sequence can be set by the user in advance, the identification can be performed once according to the target stimulation frequency sequence. Specifically, the user to be detected inputs a stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ], when the stimulation frequency sequence [f' 1 , f' 2 , f' 3 ,...
  • f' n is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ], indicating that the identification of the user to be detected is passed, and the identification of the SSVEP signal can be continued;
  • the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is different from the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ], the to-be-detected User identification failed and identity failed.
  • the subsequent identification of the SSVEP signal may still be continued, but the identification result is an identity error and cannot be passed; or, the subsequent Regarding the identification of the SSVEP signal, the identification failure is directly determined.
  • each frequency in the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] may be arranged in order or randomly arranged, therefore, in determining the stimulation frequency sequence [f' 1 , Whether f' 2 , f' 3 , ..., f' n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ] may include determining each frequency of the user input to be detected Whether the order of the corresponding frequencies in the target stimulation frequency sequence is the same, whether the magnitude of each frequency input is the same as the magnitude of the corresponding frequency in the target stimulation frequency sequence, and whether the number of input frequencies is the same as the number of frequencies in the target stimulation frequency sequence.
  • the user when the user initializes the identification system, the user selects [12, 8, 20] Hz as the target stimulation sequence, and the real brain electrical signal X(12, t), X(8, t) under three frequency stimulations. ) and X (20, t) are saved to the safe identification system.
  • the system requests the user to be detected to input the stimulation frequency sequence instruction to end with the # key. If the user to be detected inputs a sequence of non-set frequency sizes such as [6, 12, 15#], or enter a sequence of [12, 20, 8#], etc., or enter [12, 8#], etc. The sequence indicates that the authentication of the user to be detected fails, and the identification is ended. Only when the user correctly inputs [12, 8, 20#], it indicates that the identity of the user to be detected this time regarding the stimulation frequency sequence input is successful, and the identity verification about the SSVEP signal can be continued.
  • the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is also possible to input the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] as the target stimulus in the user to be detected regardless of all the factors in the order, the size and the number.
  • the user input stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f can also be determined.
  • ' n ' is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ].
  • the user to be detected inputs the stimulation frequency sequence [f' 1 , f′ 2 , f′ 3 , . . . , f′ n ], and determines the stimulation frequency sequence [f′ 1 , f′ 2 , f′ 3 , ..., f' n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ], regardless of the input order, ie the size of each frequency input and the target stimulation sequence
  • the stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is determined with the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ] are the same.
  • the user when the user initializes the identification system, the user selects [12, 8, 20] Hz as the target stimulation sequence, and the real brain electrical signals X (12, t), X (8, t) and X (20, t) are saved to the safe identification system.
  • the system requests the user to be detected to input the stimulation frequency sequence instruction to end with the # key. If the user to be detected inputs a sequence of non-set frequency sizes such as [6, 12, 15#], or enters a sequence with an incorrect number such as [12, 8#], it indicates that the identity verification of the user to be detected fails, and the identification is ended.
  • [12,20,8#], etc. is only a sequence of sequence errors, which can indicate that the user to be detected this time is successful in the authentication of the stimulus frequency sequence input, and can continue to perform identity verification on the SSVEP signal.
  • the user to be detected inputs the stimulation frequency sequence [f' 1 , f′ 2 , f′ 3 , . . . , f′ n ], and determines the stimulation frequency sequence [f′ 1 , f′ 2 , f′ 3 , ..., if f' n ] is the same as the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ], the number of inputs, that is, the size and target of each frequency input, may be disregarded.
  • the stimulation frequency sequence can be determined when the corresponding frequency in the stimulation sequence is the same and the order of the input frequency is the same as the frequency sequence in the target stimulation sequence, or only if the magnitude of each frequency input is the same as the corresponding frequency in the target stimulation sequence.
  • f' 1 , f' 2 , f' 3 , ..., f' n ] are identical to the target stimulation sequence [f 1 , f 2 , f 3 , ..., f n ].
  • the user can set a large number of frequencies, for example, setting 10 frequencies to form an initial stimulation frequency sequence.
  • only a stimulation frequency sequence composed of a small number of frequencies may be input. If the frequency included in the stimulation frequency sequence belongs to the initial stimulation frequency sequence set by the user, the identification is passed, otherwise it does not pass.
  • the user sets an initial stimulation frequency sequence composed of 10 frequencies, and when the user to be detected performs identification, the input stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is [12 , 8, 20], if the initial stimulation frequency sequence is found and it is determined that the three frequencies are included, the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] is [12, 8, 20 ], the target stimulation sequence is a subset of 10 frequency sets set by the user, the [f' 1 , f' 2 , f' 3 , ..., f' n ] and [f 1 , f 2 , f 3 , ..., f n ] is the same, the identification is passed, and the identification of the SSVEP signal is continued; if the initial stimulation frequency sequence is searched, it is determined that the three frequencies are not included or only one or two frequencies are included, for example, only The target stimulation frequency
  • the user sets an initial stimulation frequency sequence composed of 10 frequencies, and when the user to be detected performs identification, the input stimulation frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is [ 12, 8, 20], if the initial stimulation frequency sequence is searched to determine that the three frequencies are consecutive, and the three frequencies are in the same order, that is, there is a frequency sequence among the 10 frequencies, that is, the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] are [12, 8, 20], the [f' 1 , f' 2 , f' 3 , ..., f' n ] and [f 1 , f 2 , f 3 , ..., f n ] are the same, then the identification is passed, and the identification of the SSVEP signal is continued; if the initial stimulation frequency
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] is [12, 8], excluding the frequency 20 Hz, or there is a target stimulation frequency sequence.
  • [12,20,8] that is, the input order is different, or the target stimulation frequency sequence is [12, 11, 20, 8], that is, the input order is discontinuous, then the [f' 1 , f' 2 is indicated . f ' 3 , . . . , f′ n ] is different from [f 1 , f 2 , f 3 , . . . , f n ], and the identification does not pass, and the identification is ended.
  • the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] may also not be used as an identification basis, and when the user to be detected needs to perform identity recognition, the identity recognition system Automatically acquiring a target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] set in advance by the user, or randomly acquiring or sequentially acquiring a part of a series of frequencies set in advance by the user to constitute a target stimulation
  • the frequency sequence [f 1 , f 2 , f 3 , ..., f n ] to be detected by the acquired target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] for performing SSVEP Identification of the signal.
  • the target i-stimulus frequency sequence [f 1 , f 2 , f 3 , ..., f n ] has an i-th frequency f i , i taken as 1, 2, 3...n.
  • the user to be detected when the user to be detected performs identity recognition, after determining the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], according to the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . .
  • f i sequentially displays n segments of corresponding stimulation signals for the user to be detected, and the display frequency of the i-th segment stimulation signals in the n-segment stimulation signals is the target stimulation frequency sequence [f 1 , f 2 ,
  • the i-th frequency f i , i in f 3 , ..., f n ] takes 1, 2, 3 , ..., n in order.
  • the first segment of the stimulation signal is displayed at a frequency of f 1 for a period of time, and then the second segment of the stimulation signal is displayed for a period of time at a frequency of f 2 , and the third segment of the stimulation signal is displayed for a period of time at a frequency of f 3 until the nth
  • the segment stimulation signal is displayed at a frequency of f n for a period of time, and then a corresponding n-segment SSVEP signal generated by the user to be detected according to the n-segment stimulation signal may be acquired.
  • the user when setting the preset SSVEP signal, the user also generates n segments according to sequentially displaying the n-segment stimulation signals corresponding to the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ].
  • the SSVEP signal is used to collect the n-segment SSVEP signal as a preset SSVEP signal, so as to detect whether the SSVEP signal of the user to be detected is correct according to the preset SSVEP signal, thereby performing identity verification.
  • the display frequency of the stimulation signal displayed for the user to be detected or the original user who sets the SSVEP signal is f i , where the stimulation signal may be a picture with a certain pattern, and the picture is displayed according to the corresponding frequency f i ;
  • the stimulation signal may also be a light, and the frequency of the switch or the light and dark of the light is the display frequency f i ; or the stimulus signal may be any other form that can induce the SSVEP signal generated by the human brain, and is displayed to the user through the frequency f i In order to obtain the SSVEP signal generated by the user to be detected or the original user's reaction to the stimulation signal.
  • the type of stimulation signal used by the original user who sets the preset SSVEP signal is consistent with the type of stimulation signal used by the user to be detected, so that errors due to different stimulation signals can be avoided.
  • the user when collecting the SSVEP signal of the original user or the user to be detected that sets the preset SSVEP signal, the user may be provided with a wearable dry electrode cap, the electrode cap including at least one dry electrode placed at the scalp of the user's pillow region, so that Acquire the user's SSVEP signal.
  • the user looks at the stimulation signal with the eye and maintains the attention.
  • the SSVEP signal induced by the user's cerebral occipital region can be acquired by the wearable EEG device (ie, the wearable dry electrode cap).
  • the collected SSVEP signals can be wirelessly transmitted to the corresponding signal analysis area for subsequent analysis and processing.
  • the synchronization information may be sent to the corresponding module that collects the SSVEP signal, so that when the SSVEP signal is acquired, each segment of the stimulation signal corresponding to the different frequency may be intercepted according to the synchronization signal.
  • the SSVEP signal that is, the n-segment stimulation signal corresponds to the n-segment SSVEP signal, and the n-segment SSVEP signal can be separately intercepted by the synchronization signal.
  • each SSVEP signal corresponds to a frequency of a stimulation signal, that is, the i-th SSVEP signal can be represented as X i (f i , T i ), and f i represents the The i-th SSVEP signal is obtained according to the stimulation signal whose display frequency is f i , that is, the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] corresponds to the SSVEP signal X 1 of the n-segment user to be detected. (f 1 , T 1 ), X 2 (f 2 , T 2 ) ...
  • the preset SSVEP signal set by the user in advance may be expressed as X' i (f i , T i ), that is, the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] corresponds to n segments of pre- Let SSVEP signals X' 1 (f 1 , T 1 ), X' 2 (f 2 , T 2 ) ... X' n (f n , T n ).
  • X i (f i , T i ) or X' i (f i , T i ) may be a matrix.
  • X i (f i , T i ) or X′ i (f i , T i ) can be reduced to one-dimensional by classical data dimension reduction methods such as principal component analysis (PCA) or canonical correlation analysis (CCA).
  • PCA principal component analysis
  • CCA canonical correlation analysis
  • Time domain signal is treated as one-dimensional time domain signals as an example.
  • the collected SSVEP signal of the user to be detected is compared with the preset SSVEP signal to determine whether the identity of the user to be detected is correct.
  • the identity recognition result may be determined by calculating a similarity between the n-segment SSVEP signal of the user to be detected and the n-segment preset SSVEP signal.
  • the identity may be identified; when the n-segment SSVEP signal and the n-segment preset SSVEP signal are When the similarity is less than the threshold, the identity of the user to be detected is incorrect, and the identity is not passed.
  • the identification can be passed, that is, the insurance can be opened.
  • the cabinet however, when the identification does not pass, the safe cannot be opened, and further, when the identification fails, the alarm device can also be activated.
  • the similarity between the n-segment SSVEP signal and the n-segment SSVEP signal can be calculated.
  • the correlation coefficient between the n-segment SSVEP signal and the n-segment preset SSVEP signal may be calculated by the time domain angle. Specifically, a time domain correlation coefficient r X (f i ) of the i-th SSVEP signal X i (f i , T i ) and the i-th segment preset SSVEP signal X′ i (f i , T i ) may be calculated,
  • the correlation coefficient can be a linear correlation coefficient, for example, a Pearson linear correlation coefficient.
  • the Pearson linear correlation coefficient can be used to measure the degree of similarity between the input signal and the real user signal. Even at the same stimulation frequency, different people have different differences in the physiological structure of the brain, for example, the visual pathway delay is different, the spectral response of the visual system is different, and there is a big difference between the SSVEP signals induced by them. Therefore, the correlation coefficient will be larger only when the input SSVEP signal is the real user's own EEG signal.
  • the transform domain correlation coefficient of the n-segment SSVEP signal and the n-segment preset SSVEP signal may also be calculated by transforming the domain angle.
  • the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal may be transformed to obtain Z i (f i , Y i );
  • the X′ i in the n-segment preset SSVEP signal ( f i , T i ) is transformed to obtain Z′ i (f i , Y i );
  • the correlation coefficient of the transform domain determined to determine Z i (f i , Y i ) and Z′ i (f i , Y i ) is r Y (f i ), i takes 1, 2, 3, . . . n; determines the correlation coefficient vector according to the transform domain correlation coefficient r Y (f i )
  • a transform domain refers to other domains that are different from the time domain (T domain), for example, a frequency domain (ie, an F domain), an S domain, a Z domain, and the like.
  • the time domain SSVEP signal X i (f i , T i ) can be transformed into the transform domain by an algorithm to obtain Z i (f i , Y i ); for example, the SSVEP signal X i (f i can be obtained by the Fourier algorithm ) , T i ) transforms into the frequency domain, and obtains Z i (f i , F i ).
  • the SSVEP signal X i (f i , t) can also be transformed into the frequency domain by the fast Fourier algorithm to obtain another Z i (f i , F i );
  • the SSVEP signal X i (f i , t) is transformed into the S domain by the Laplace algorithm to obtain Z i (f i , S i ).
  • the transform domain correlation coefficient r Y (f i ) may include a frequency domain correlation coefficient r F (f i ), that is, a frequency domain correlation coefficient of the n-segment SSVEP signal and the n-segment preset SSVEP signal may be calculated by a frequency domain angle.
  • the similarity of the amplitude-frequency response is calculated as an example.
  • the i-th segment SSVEP signal X i (f i , T i ) is Fourier transformed to obtain a corresponding amplitude-frequency response signal Z i (f i , F i ), and similarly, the i-th segment is pre-
  • the SSVEP signal X' i (f i , T i ) also perform Fourier transform to obtain Z′ i (f i , F i ), and calculate Z i (f i , F i ) and Z′ i (f i ,
  • the frequency domain correlation coefficient of F i ) is r F (f i ).
  • the similarity coefficient of the phase-frequency response of the n-segment SSVEP signal and the n-segment preset SSVEP signal may also be calculated, and the embodiment of the present invention is not limited thereto.
  • the time domain correlation coefficient r X (f i ) and the transform domain correlation coefficient determined by the above two methods are r Y (f i ), and the n segment SSVEP signal and the n segment preset SSVEP are determined.
  • the correlation coefficient vector may be determined according to the n time domain correlation coefficients r X (f i ) and the n transform domain correlation coefficients being r Y (f i ) Correlation coefficient vector
  • the elements in the time may represent time domain correlation coefficients, and may also represent transform domain correlation coefficients, wherein the transform domain correlation coefficients may be frequency domain correlation coefficients, or may be other correlation coefficients.
  • the correlation coefficient vector The n elements may be n, and the n elements may be n time domain correlation coefficients r X (f i ), or the n transform domain correlation coefficients may be r Y (f i ), or the n elements may include i time domain correlation coefficients r X (f i ), and (ni) transform domain correlation coefficients corresponding to the remaining (ni) time domain correlation coefficients r X (f i ) are r Y (f i ); Correlation coefficient vector It is also possible to include 2n elements including n time domain correlation coefficients r X (f i ) and n transform domain correlation coefficients r y (f i ), and the present invention is not limited thereto.
  • the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal is determined. Specifically, you can pass the formula Determining the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal, wherein Representing a weight parameter vector, that is, the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal is equal to the weight parameter vector Transpose and correlation coefficient vector Inner product.
  • the weight parameter vector Each element in the corresponding coefficient vector The weight of the correlation coefficient represented by each element.
  • the weight parameter vector Each element in the set is set equal to among them, Representation correlation coefficient vector The number of elements included in the. Or when the statistics show the correlation coefficient vector When an element in the middle can better reflect the difference between the real user and the non-real user compared to other elements, the weight value corresponding to the element can be increased.
  • the element herein can better reflect the difference between the real user and the non-real user, and the difference between the non-real user and the original real user is larger than the other elements.
  • the original user may perform multiple acquisitions. After determining the preset SSVEP signal, each of the original user's SSVEP signal and the preset SSVEP signal may determine a correlation coefficient vector. Multiple acquisitions to determine multiple sets of correlation coefficient vectors Multiple sets of correlation coefficient vectors The average value can be used as the original user's original correlation coefficient vector It can reflect the similarity between the original user itself and each preset SSVEP signal, ie.
  • the correlation coefficient vector An element and the original correlation coefficient vector The difference between the corresponding elements is greater than the other elements and the original correlation coefficient vector The difference between the corresponding elements indicates that the element can strongly reflect the distinguishing between the original user and the user to be detected, so that the weight value corresponding to the element can be modulated to be too large.
  • the target stimulation frequency sequence is [12, 8], and the corresponding time domain correlation coefficients r(12) and r(8) are calculated, so that the correlation coefficient vector is obtained. Only the two correlation coefficients, r(12) and r(8), can be set.
  • the weight value may be modified.
  • determining that the r(12) difference is large means that, for example, when the original user sets the preset SSVEP signal, the determined R(12) and R(8) are both 0.9, but the calculated user to be detected r(8) ) is 0.5, r(12) is 0.2, and the difference between r(12) and R(12) is greater than 0.7, and the difference between r(8) and R(8) is 0.4, then the corresponding r(12) can be increased.
  • the r(12) of the user to be detected is 0.2, but when the original user sets the preset SSVEP signal, the determined R(12) is 0.8, and the r(8) of the user to be detected is 0.3, the original When the user sets the preset SSVEP signal, the determined R(8) is 0.9. At this time, the r(12) and R(12) of the user to be detected are different by 0.6, and the difference between r(8) and R(8) is also 0.6. Therefore, the weight value of the r(12) of the user to be detected may not be increased.
  • the identity is determined according to the similarity between the n-segment SSVEP signal of the user to be detected and the n-segment preset SSVEP signal.
  • the similarity is greater than or equal to the threshold, it indicates that the identity of the user to be detected is correct, that is, the identification based on the SSVEP signal is used; when the similarity is less than the threshold, the identity of the user to be detected is incorrect, that is, cannot pass this time.
  • Identification based on SSVEP signals may be set according to actual conditions, and the threshold may be continuously updated according to multiple tests.
  • the size of the sequence numbers of the above processes does not mean the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not be taken to the embodiments of the present invention.
  • the implementation process constitutes any limitation.
  • the method for identifying the EEG signal based on the embodiment of the present invention sets the target stimulation frequency sequence to display the stimulation signal for the user to be detected according to a certain frequency, thereby collecting the SSVEP signal generated by the user to be detected, and the SSVEP signal is
  • the preset SSVEP signals are compared for identification of the user to be detected. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • identification can also be performed, so that the exclusive stimulation sequence plus the SSVEP signal dual authentication mode improves the level of device confidentiality and is less likely to be invaded.
  • an apparatus for identifying an EEG-based identity includes:
  • a determining unit 210 configured to determine a target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], where n is a positive integer;
  • the display unit 220 is configured to display, for the user to be detected, an n-segment stimulation signal corresponding to the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . , f n ], and the i-th segment stimulation signal in the n-segment stimulation signal
  • the display frequency is the i-th frequency f i , i of the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ], 1 , 2 , 3 , ... n;
  • the acquiring unit 230 is configured to acquire an n-stage steady-state visual evoked potential SSVEP signal generated by the user to be detected due to the n-segment stimulation signal;
  • the processing unit 240 is configured to determine that the identity of the user to be detected is correct when the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal is greater than or equal to a threshold; and when the n-segment SSVEP signal and the n-segment are When the similarity of the SSVEP signal is less than the threshold, it is determined that the identity of the user to be detected is incorrect.
  • the apparatus for identifying an EEG based signal displays a stimulus signal for the user to be detected according to the target stimulation frequency sequence, thereby collecting an SSVEP signal generated by the user to be detected, and performing the SSVEP signal with the preset SSVEP signal.
  • the identification of the user to be detected is performed. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • the determining unit 210 is specifically configured to: acquire a sequence of the to-be-detected stimulation frequency [f′ 1 , f′ 2 , f′ 3 , . . . , f′ n ] input by the user to be detected; when the stimulus to be detected When the frequency sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is the same as the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ], the target is acquired The n-segment stimulation signal corresponding to the stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ]; when the stimulation frequency sequence to be detected [f' 1 , f' 2 , f' 3 , ..., When f′ n ] is different from the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . . ,
  • the processing unit 240 is specifically configured to: determine the n-segment SSVEP signal and the n-segment preset SSVEP according to the n-segment SSVEP signal of the at least one domain and the n-segment preset SSVEP signal of the corresponding at least one domain.
  • Each element in the representation represents a correlation coefficient between the n-segment SSVEP signal and the n-segment preset SSVEP signal; Determining the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal, wherein Represents a weight parameter vector, which is a weight parameter vector Each element in the representation represents the correlation coefficient vector The weight value of the corresponding element in .
  • the processing unit 240 is specifically configured to: determine an i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal and X′ i (f i in the n-segment preset SSVEP signal,
  • the time domain correlation coefficient of T i ) is r X (f i ), i is 1, 2, 3...n; and the correlation coefficient vector is determined according to the time domain correlation coefficient r X (f i )
  • the processing unit 240 is specifically configured to: transform the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , Y i ); X' i (f i , T i ) in the preset SSVEP signal is transformed to obtain Z′ i (f i , Y i ); determining Z i (f i , Y i ) and Z′ i (f i , Y i
  • the transform domain correlation coefficient is r Y (f i ), i is 1, 2, 3...n; the correlation coefficient vector is determined according to the transform domain correlation coefficient r Y (f i )
  • the processing unit 240 is specifically configured to: perform Fourier transform on the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , F i ); X' i (f i , T i ) in the n-stage preset SSVEP signal is subjected to Fourier transform to obtain Z′ i (f i , F i ); determining Z i (f i , F i ) and Z′ i
  • the frequency domain correlation coefficient of (f i , F i ) is r F (f i ), i is 1, 2, 3...n; and the correlation coefficient vector is determined according to the frequency domain correlation coefficient r F (f i )
  • the apparatus 200 based on EEG-based identification may correspond to performing the method 100 in the embodiments of the present invention, and the above and other operations and/or functions of the respective modules in the apparatus 200 are respectively Corresponding processes for implementing the various methods in FIG. 1 are omitted here for brevity.
  • the apparatus for identifying an EEG based signal displays a stimulus signal for the user to be detected according to the target stimulation frequency sequence, thereby collecting an SSVEP signal generated by the user to be detected, and performing the SSVEP signal with the preset SSVEP signal.
  • the identification of the user to be detected is performed. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • identification can also be performed, so that the exclusive stimulation sequence plus the SSVEP signal dual authentication mode improves the level of device confidentiality and is less likely to be invaded.
  • an embodiment of the present invention further provides an apparatus 300 for identity recognition based on an EEG signal, including a processor 310 and a memory 320, and may further include a bus system 330.
  • the processor 310 and the memory 320 are connected by a bus system 330 for storing instructions for executing instructions stored by the memory 320.
  • the memory 320 stores the program code, and the processor 310 can call the program code stored in the memory 320 to perform the following operations: determining the target stimulation frequency sequence [f 1 , f 2 , f 3 , . . .
  • the i-th frequency f i , i of the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ] is 1 , 2 , 3 , ...
  • n obtaining the user to be detected due to the n-segment stimulation signal
  • the generated n-stage steady-state visual evoked potential SSVEP signal when the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal is greater than or equal to the threshold, determining that the identity of the user to be detected is correct; when the n-segment SSVEP signal and the When the similarity of the n-segment preset SSVEP signal is less than the threshold, it is determined that the identity of the user to be detected is incorrect.
  • the apparatus for identifying an EEG based signal displays a stimulus signal for the user to be detected according to the target stimulation frequency sequence, thereby collecting an SSVEP signal generated by the user to be detected, and performing the SSVEP signal with the preset SSVEP signal.
  • the identification of the user to be detected is performed. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • the processor 310 may be a central processing unit ("CPU"), and the processor 310 may also be other general-purpose processors, digital Signal processor (DSP), application specific integrated circuit (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the memory 320 can include read only memory and random access memory and provides instructions and data to the processor 310. A portion of the memory 320 may also include a non-volatile random access memory. For example, the memory 320 can also store information of the device type.
  • the bus system 330 may include a power bus, a control bus, a status signal bus, and the like in addition to the data bus. However, for clarity of description, various buses are labeled as bus system 330 in the figure.
  • each step of the foregoing method may be completed by an integrated logic circuit of hardware in the processor 310 or an instruction in a form of software.
  • the steps of the method disclosed in the embodiments of the present invention may be directly implemented as a hardware processor, or may be performed by a combination of hardware and software modules in the processor.
  • the software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
  • the storage medium is located in the memory 320, and the processor 310 reads the information in the memory 320 and combines the hardware to perform the steps of the above method. To avoid repetition, it will not be described in detail here.
  • the processor 310 is configured to: acquire a sequence of to-be-detected stimulation frequencies [f′ 1 , f′ 2 , f′ 3 , . . . , f′ n ] input by the user to be detected; when the stimulation frequency to be detected is When the sequence [f' 1 , f' 2 , f' 3 , ..., f' n ] is the same as the target stimulation frequency sequence [f 1 , f 2 , f 3 , ..., f n ], the target stimulus is acquired.
  • the processor 310 is configured to: determine the n-segment SSVEP signal and the n-segment preset SSVEP signal according to the n-segment SSVEP signal of the at least one domain and the n-segment preset SSVEP signal of the corresponding at least one domain.
  • Correlation coefficient vector Correlation coefficient vector
  • Each element in the representation represents a correlation coefficient between the n-segment SSVEP signal and the n-segment preset SSVEP signal; Determining the similarity between the n-segment SSVEP signal and the n-segment preset SSVEP signal, wherein Represents a weight parameter vector, which is a weight parameter vector
  • Each element in the representation represents the correlation coefficient vector The weight value of the corresponding element in .
  • the processor 310 is configured to: determine an i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal and X′ i (f i , T in the n-segment preset SSVEP signal) i) a time domain correlation coefficient r X (f i), i taking 1,2,3 ...... n; r X (f i), the correlation coefficient vector is determined based on the time domain correlation coefficient
  • the processor 310 is configured to: transform the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , Y i ); Let X' i (f i , T i ) in the SSVEP signal be transformed to obtain Z′ i (f i , Y i ); determine Z i (f i , Y i ) and Z′ i (f i , Y i ) The transform domain correlation coefficient is r Y (f i ), i is 1, 2, 3...n; the correlation coefficient vector is determined according to the transform domain correlation coefficient r Y (f i )
  • the processor 310 is configured to perform Fourier transform on the i-th SSVEP signal X i (f i , T i ) in the n-segment SSVEP signal to obtain Z i (f i , F i ); X' i (f i , T i ) in the n-segment preset SSVEP signal is subjected to Fourier transform to obtain Z′ i (f i , F i ); determining Z i (f i , F i ) and Z′ i ( The frequency domain correlation coefficient of f i , F i ) is r F (f i ), i is 1, 2, 3...n; and the correlation coefficient vector is determined according to the frequency domain correlation coefficient r F (f i )
  • the EEG-based identity recognition device 300 may correspond to the EEG-based identity-based device 200 in the embodiment of the present invention, and may correspond to performing an embodiment according to the present invention.
  • the above-mentioned and other operations and/or functions of the respective modules in the method 100 are respectively omitted in order to implement the respective processes of the respective methods in FIG. 1 for brevity.
  • the apparatus for identifying an EEG based signal displays a stimulus signal for the user to be detected according to the target stimulation frequency sequence, thereby collecting an SSVEP signal generated by the user to be detected, and performing the SSVEP signal with the preset SSVEP signal.
  • the identification of the user to be detected is performed. Due to the high signal-to-noise ratio of the SSVEP signal, the signal can be more easily detected.
  • the stimulation duration can be shortened; the SSVEP signal is mainly concentrated in the human occipital region, requiring only a few electrodes, such as An electrode can collect a rich amount of information signals, which is convenient to use; SSVEP signal is a primary visual cortex-inducing signal, which does not require the participation of human advanced cognitive activities, so it is less affected by human mental state, and signal characteristics Relatively more stable; the amplitude frequency response and phase frequency response of the SSVEP signal can directly correspond to the system physiological characteristics of the human primary visual cortex, making the identification system based on this more confidential and difficult to be copied and forged.
  • identification can also be performed, so that the exclusive stimulation sequence plus the SSVEP signal dual authentication mode improves the level of device confidentiality and is less likely to be invaded.
  • the disclosed systems, devices, and methods may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs. The purpose of implementing the solution of this embodiment is achieved.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product.
  • the technical solution of the present invention which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including
  • the instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like. .

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)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
  • Electrotherapy Devices (AREA)

Abstract

基于脑电信号的身份识别的方法（100）和装置（200）。该方法（100）包括：确定目标刺激频率序列（S110）；为待检测用户显示该目标刺激频率序列对应的n段刺激信号（S120）；获取该待检测用户由于该n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号（S130）；当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定该待检测用户身份正确；否则，确定该待检测用户身份错误（S140）。基于脑电信号的身份识别的方法（100）和装置（200），通过SSVEP信号进行身份识别，相比于现有脑电身份识别方法可以缩短刺激时长，信号特征相对更稳定，基于此构建的身份识别系统更加的保密，不易被复制伪造。

Description

基于脑电信号的身份识别的方法和装置 技术领域
本发明涉及信息技术领域，基于脑电信号的身份识别的方法和装置。
背景技术
进入万物互联互通的时代，由于连接的智能设备数目的指数级增长，信息安全问题变得尤为重要。在信息安全领域，身份识别是一项被广泛应用的技术。随着信息安全问题越来越重要，各种身份识别方法的研究也正在成为研究热点。
传统的身份识别方法一般借助钥匙、身份证件、用户名和密码等物品或信息口令来实现。然而，这种传统的身份识别方法极易被盗用、丢失或遗忘，已经不能够完全满足人们的期望和要求。于是人们开始诉诸于利用生物特征来进行身份识别。生物特征识别技术通常指利用人体固有的一些生理特性或行为特征来进行身份识别。人体的生理特性一般包括：人脸、指纹、掌形、虹膜等；人的行为特征可包括：笔记、步态等。
然而，目前已有的生物特征识别技术也面临着一些问题。例如，对于通过指纹进行身份识别，用明胶制成的假手指就可以顺利地骗过指纹识别系统；对于通过虹膜进行身份识别，在隐形眼镜上蚀刻出的虚假虹膜特征可以让虹膜识别系统无法辨别真假。这些问题对生物特征识别技术提出了新的挑战，也启发人们不断探索新的生物特征识别方法。
近年来，人们开始考虑将脑电作为一种新型的生物特征应用在身份识别当中。研究表明，即便在同样的外部刺激下或者人们在思考同样的问题时，不同主体的大脑所诱发产生的脑电信号也是不同的，即脑电具有显著的个体差异性。与此同时，脑电具有难以复制和伪造、可受主体自主注意力调制等众多优势。因此，目前出现了一系列基于各种模式脑电的身份识别方法。如基于静息脑电的身份识别方法、基于想象运动状态脑电特征的身份识别方法、基于P300事件相关电位的脑电身份识别方法等等。
但是上述所用脑电信号信噪比一般比较低，信号特征不够稳定，通常需要采集多个导联的脑电信号且需要大量的训练样本，使用起来不够方便。例如，基于静息脑电的身份识别方法，静息状态的自发脑电具有高度的非平稳性，也容易受个体状态的影响，个体内的变动性较大；并且在测试时，用户一般需要覆盖全脑的多个电极采集脑电信号，用户使用不便。再例如，基于P300事件相关电位的脑电身份识别方法，P300事件相关电位的信噪比很低，需要大量的重复刺激并叠加平均后才可以获得稳定的波形，认证时长较长，使用不便；并且，P300事件相关电位牵扯到人的高级认知作用，多由新奇刺激诱发，易受用户主体的精神状态影响，长时间的刺激会让用户出现适应性反应，导致诱发的P300事件相关电位衰减， 不利于信号特征的提取，可靠性差。
发明内容
本申请提供了一种基于脑电信号的身份识别的方法和装置，能够根据稳态视觉诱发电位SSVEP信号进行身份识别，提高身份识别的可靠性。
第一方面，提供了一种基于脑电信号的身份识别的方法，该方法包括：确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数；为待检测用户显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，该n段刺激信号中第i段刺激信号的显示频率为该目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n；获取该待检测用户由于该n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号；当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定该待检测用户身份正确；当该n段SSVEP信号与该n段预设SSVEP信号的相似度小于该阈值时，确定该待检测用户身份错误。
因此，本申请的基于脑电信号的身份识别的方法，根据目标刺激频率序列为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。
应理解，采用SSVEP信号进行身份识别，该SSVEP信号是由稳定的周期性的视觉刺激诱发人脑产生，也就是无论是原始用户事先录入的预设SSVEP信号，还是待检测用户在身份识别时被采集的SSVEP信号，都需要稳定的周期性视觉刺激来诱发人脑产生SSVEP信号。因此，需要确定至少一个视觉刺激的频率。
应理解，采集SSVEP信号时，可以让用户带上可穿戴干电极帽，该电极帽至少包括一个置于用户枕区头皮处的干电极，以便于采集用户的SSVEP信号。在采集用户的SSVEP信号时，用户用眼睛注视刺激信号并保持注意力。此时，用户大脑枕区诱发的SSVEP信号便可通过可穿戴脑电设备(即可穿戴干电极帽)采集到。采集到的SSVEP信号可被无线传输到相应的信号分析区域，供后续分析处理识别。
结合第一方面，在第一方面的一种实现方式中，该确定目标刺激频率序列[f1,f2,f3,……,fn]，包括：获取该待检测用户输入的待检测刺激频率序列[f′1,f′2,f′3,……,f′n]；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]相同时，获取该目标刺激频率序列[f1,f2,f3,……,fn]对应的 该n段刺激信号；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]不同时，确定该待检测用户身份错误。
因此，对于设置的目标刺激频率序列，进行身份识别，这样，专属的刺激序列加SSVEP信号双认证的模式，提高了设备保密的等级，更不易被侵入。
可选地，用户设置的该目标刺激频率序列[f1,f2,f3,……,fn]中每个频率的范围一般在6Hz至100Hz之间，且该n个频率可以按照大小顺序排列，或者随机排列，该n个频率中可以包括至少两个相同频率。
可选地，由于该目标刺激频率序列[f1,f2,f3,……,fn]中每个频率可以按照顺序排列或者随机排列，因此，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同时，可以包括确定两个序列中对应地每个频率的顺序相同、对应的每个频率的大小相同以及两个序列中频率的个数相同。
可选地，还可以不考虑上述关于顺序、大小和个数中全部因素，在该待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]为目标刺激序列[f1,f2,f3,……,fn]的子集时，也可以确定该待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。
例如，待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]是否相同时，可以不考虑输入顺序，即在输入的每个频率的大小与目标刺激序列中对应频率大小相同以及输入频率的个数与目标刺激序列也相同时，确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。
再例如，，待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]是否相同时，还可以不考虑输入个数，即在输入的每个频率的大小与目标刺激序列中对应频率大小相同以及输入频率的顺序与目标刺激序列中频率顺序也相同时，或者仅在输入的每个频率的大小与目标刺激序列中对应频率大小相同，均可确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。
应理解，该目标刺激频率序列[f1,f2,f3,……,fn]也可以不作为身份识别依据，当待检测用户需要进行身份识别时，由身份识别装置自动获取事先由用户设置的目标刺激频率序列[f1,f2,f3,……,fn]，或者随机获取或按顺序获取事先由用户设置的一系列频率中部分频率，构成目标刺激频率序列[f1,f2,f3,……,fn]，通过获取到的该目标刺激频率序列[f1,f2,f3,……,fn]对待检测用户显示刺激信号，并采集待检测用户因该刺激信号产生的SSVEP信号，进行SSVEP信号的身份识别。
结合第一方面及其上述实现方式，在第一方面的另一种实现方式中，为待检测用户显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，包括：为待检测用户显示n段刺激信号，其中，第一段刺激信号以f1为频率显示一段时间，第二段刺激信号以f2为频率显示一段时间，第三段刺激信号以f3为频率显示一段 时间，一直到第n段刺激信号以fn为频率显示一段时间。
应理解，为待检测用户显示上述n段刺激信号后，则可以采集到待检测用户根据该n段刺激信号产生的对应的n段SSVEP信号。同样的，在用户设置预设SSVEP信号时，也是根据依次显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，产生n段SSVEP信号，采集该n段SSVEP信号作为预设SSVEP信号进行保存。
应理解，该刺激信号可以为带有一定图案的图片，该图片按照对应的频率fi显示；或者该刺激信号也可以为灯光，灯光的开关或明暗交替的频率即为显示频率fi；或者，该刺激信号还可以为其他任何可以诱发人脑产生SSVEP信号的形式，通过频率fi向用户显示，以便于获取该待检测用户对于刺激信号的反应产生的SSVEP信号。
结合第一方面及其上述实现方式，在第一方面的另一种实现方式中，该方法还包括：根据至少一个域的该n段SSVEP信号与对应的至少一个域的该n段预设SSVEP信号，确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000001
该相关系数向量
Figure PCTCN2016104445-appb-000002
中的每个元素表示该n段SSVEP信号与该n段预设SSVEP信号的相关系数；将
Figure PCTCN2016104445-appb-000003
确定为该n段SSVEP信号与该n段预设SSVEP信号的相似度，其中，
Figure PCTCN2016104445-appb-000004
表示权重参数向量，该权重参数向量
Figure PCTCN2016104445-appb-000005
中的每个元素表示该相关系数向量
Figure PCTCN2016104445-appb-000006
中对应元素的权重值。
具体地，该权重参数向量
Figure PCTCN2016104445-appb-000007
中每个元素，对应于相关系数向量
Figure PCTCN2016104445-appb-000008
中每个元素表示的相关系数的权重。例如，该权重参数向量
Figure PCTCN2016104445-appb-000009
中每个元素均设置为等于
Figure PCTCN2016104445-appb-000010
其中，
Figure PCTCN2016104445-appb-000011
表示相关系数向量
Figure PCTCN2016104445-appb-000012
中包括的元素的个数。或者，当该相关系数向量
Figure PCTCN2016104445-appb-000013
中某些元素相对较小时，说明该元素可以较强的反映了的真实用户与非真实用户的区分性，因此可以调制该元素对应的权重值偏大。
应理解，根据该待检测用户的n段SSVEP信号与n段预设SSVEP信号的相似度，进行身份识别。当该相似度大于或等于阈值时，说明该待检测用户身份正确，即通过本次基于SSVEP信号的身份识别；当该相似度小于阈值时，说明该待检测用户身份错误，即不能通过本次基于SSVEP信号的身份识别。可选地，该阈值可以根据实际情况进行设置，并且可以根据多次测试，不断更新该阈值。
结合第一方面及其上述实现方式，在第一方面的另一种实现方式中，该确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000014
包括：确定该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)与该n段预设SSVEP信号中的X′i(fi,Ti)的时域相关系数为rX(fi)，i取1、2、3……n；根据该时域相关系数rX(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000015
结合第一方面及其上述实现方式，在第一方面的另一种实现方式中，该确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000016
包括：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；将该n段预设 SSVEP信号中的X′i(fi,Ti)进行变换得到Z′i(fi,Yi)；确定Zi(fi,Yi)与Z′i(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；根据该变换域相关系数rY(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000017
结合第一方面及其上述实现方式，在第一方面的另一种实现方式中，该确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000018
包括：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换得到Zi(fi,Fi)；将该n段预设SSVEP信号中的X′i(fi,Ti)进行傅里叶变换得到Z′i(fi,Fi)；确定Zi(fi,Fi)与Z′i(fi,Fi)的频域相关系数为rF(fi)，i取1、2、3……n；根据该频域相关系数rF(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000019
应理解，当采集到的SSVEP信号是多个电极的信号时，Xi(fi,Ti)或者X′i(fi,Ti)可以为一个矩阵。此时，可以用主成分分析(PCA)或典型相关分析(CCA)等经典数据降维方法，将Xi(fi,Ti)或者X′i(fi,Ti)降为一维时域信号。本申请以Xi(fi,Ti)和X′i(fi,Ti)为一维时域信号为例进行说明。
应理解，该变换域相关系数为rY(fi)可以包括频域相关系数为rF(fi)。
可选地，该时域相关系数为rX(fi)和变换域相关系数为rY(fi)，均可以为线性相关系数，例如，皮尔森线性相关系数。
可选地，根据n个时域相关系数rX(fi)和/或n个变换域相关系数rY(fi)，确定相关系数向量
Figure PCTCN2016104445-appb-000020
该相关系数向量
Figure PCTCN2016104445-appb-000021
中的元素可以表示时域相关系数，也可以表示频域相关系数。例如，该相关系数向量可以包括n个元素，该n个元素可以包括i个时域相关系数rX(fi)，以及包括剩余(n-i)个时域相关系数rX(fi)对应的(n-i)个变换域相关系数为rY(fi)；再例如，该2n个元素包括n个时域相关系数rX(fi)以及n个变换域相关系数为rY(fi)。
第二方面，提供了一种基于脑电信号的身份识别的装置，用于执行上述第一方面或第一方面的任意可能的实现方式中的方法。具体地，该装置包括用于执行上述第一方面或第一方面的任意可能的实现方式中的方法的单元。
第三方面，提供了一种基于脑电信号的身份识别的装置，包括：存储单元和处理器，该存储单元用于存储指令，该处理器用于执行该存储器存储的指令，并且当该处理器执行该存储器存储的指令时，该执行使得该处理器执行第一方面或第一方面的任意可能的实现方式中的方法。
第四方面，提供了一种计算机可读介质，用于存储计算机程序，该计算机程序包括用于执行第一方面或第一方面的任意可能的实现方式中的方法的指令。
附图说明
为了更清楚地说明本发明实施例的技术方案，下面将对本发明实施例中所需要使用的附图作简单地介绍，显而易见地，下面所描述的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可 以根据这些附图获得其他的附图。
图1是根据本发明实施例的基于脑电信号的身份识别的方法的示意性流程图。
图2是根据本发明实施例的基于脑电信号的身份识别的装置的示意性框图。
图3是根据本发明另一实施例的基于脑电信号的身份识别的装置的示意性框图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明的一部分实施例，而不是全部实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例，都应属于本发明保护的范围。
图1示出了根据本发明实施例的基于脑电信号的身份识别的方法100的示意性流程图。该方法100可以由需要进行身份验证的设备执行，例如，保险柜上安装的身份识别装置可以按照该方法100进行身份验证。如图1所示，该方法100包括：
S110，确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数。
应理解，本发明实施例的基于脑电信号的身份识别的方法，需事先采集并保存用户的脑电信号作为预设脑电信号，当进行身份验证或身份识别时，采集待检测用户的脑电信号，将该脑电信号与保存的预设脑电信号进行对比，若与预设脑电信号相符，说明待检测用户身份正确，身份验证通过；若与预设脑电不相符，说明待检测用户身份错误，身份验证无法通过。
具体地，本发明实施例进行身份验证采集的脑电信号是指稳态视觉诱发电位(Steady-State Visual Evoked Potentials，SSVEP)信号，根据该SSVEP信号进行身份识别。该SSVEP信号一般由稳定的周期性的视觉刺激诱发人脑产生，一般在人的大脑的枕区可以记录到该SSVEP信号。该SSVEP信号中主要包含有与刺激信号同频率的脑电成分，同时由于视觉系统的非线性特性以及自发脑电的影响，其信号中还包含各次谐波成分以及其他频率成分。由于人的大脑生理结构上的高度复杂性和差异性，例如，自发背景脑电的差异，视觉通路的延迟差异，视觉系统的响应差异等，不同人诱发的SSVEP信号存在着很大差异。相比较于其他脑电信号，例如事件相关电位(Event Related Potentials，ERP)信号，或视觉诱发电位(Visual Evoked Potentials，VEP)信号，SSVEP信号更加稳定且信噪比更高。
在本发明实施例中，SSVEP信号是由稳定的周期性的视觉刺激诱发人脑产生，也就是无论是用户录入的预设SSVEP信号，还是待检测用户被采集的SSVEP信号，都需要稳定的周期性视觉刺激来诱发人脑产生SSVEP信号。因此，需要确定至少一个视觉刺激的频率。
具体地，对于用户在设置预设SSVEP信号时，需要先选择目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数，该目标刺激频率序列中每个频率对应于一段刺激信号的显示频率，例如，目标刺激频率序列中的频率fi指第i段刺激信号的显示频率即为fi，从而获得用户对于第i段刺激信号的预设SSVEP信号。应理解，这里的用户指身份识别时的原始用户，例如，开启保险柜时要进行身份识别，该保险柜的持有人即可以为本发明实施例中的用户，该用户会设置预设SSVEP信号作为身份识别的依据。
可选地，用户设置的该目标刺激频率序列[f1,f2,f3,……,fn]中每个频率的范围一般在6Hz至100Hz之间，且该n个频率可以按照大小顺序排列，或者随机排列，例如，用户可以设置该目标刺激频率序列为[20,57,38,64]，本发明实施例并不限于此。可选地，该n个频率中可以包括至少两个相同频率值，即该n个频率数值可以重复。
在本发明实施例中，待检测用户在需要进行身份识别时，先确定该目标刺激频率序列[f1,f2,f3,……,fn]。由于该目标刺激频率序列可由用户事先设置，因此可以根据该目标刺激频率序列先进行一次身份识别。具体地，待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]，当该刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同时，说明该待检测用户身份识别通过，可以继续进行关于SSVEP信号的身份识别；当该刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]不同时，说明该待检测用户身份识别无法通过，身份识别失败。可选地，在确定目标刺激序列时已经确定待检测用户身份识别失败时，仍然可以继续后续关于SSVEP信号的身份识别，但是识别结果为身份错误，无法通过；或者，也可以不再进行之后的关于SSVEP信号的身份识别，直接确定身份识别失败。
可选地，由于该目标刺激频率序列[f1,f2,f3,……,fn]中每个频率可以按照顺序排列或者随机排列，因此，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]是否相同时，可以包括确定待检测用户输入的每个频率与目标刺激频率序列中对应频率的顺序是否相同、输入的每个频率的大小与目标刺激频率序列中对应频率的大小是否相同以及输入频率的个数与目标刺激频率序列中频率个数是否相同。例如，用户初始化设置身份识别系统时，用户选取[12,8,20]Hz作为目标刺激序列，并将三个频率刺激下的自己真实脑电信号X(12,t)、X(8,t)与X(20,t)保存到保险箱身份识别系统中。当待检测用户请求身份认证时，系统要求待检测用户输入刺激频率序列指令以#键结束。若待检测用户输入[6,12,15#]等非设定频率大小序列，或者输入[12,20,8#]等顺序错误的序列，或输入[12,8#]等个数错误的序列，均说明待检测用户身份验证失败，结束身份识别。只有当用户正确输入[12,8,20#]时，则说明待检测用户本次关于刺激频率序列输入的身份验证成功，可以继续进行关于SSVEP信号的身份验证。
另外，还可以不考虑上述关于顺序、大小和个数中全部因素，在该待检测用 户输入刺激频率序列[f′1,f′2,f′3,……,f′n]为目标刺激序列[f1,f2,f3,……,fn]的子集时，也可以确定该待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。
可选地，待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]是否相同时，可以不考虑输入顺序，即在输入的每个频率的大小与目标刺激序列中对应频率大小相同以及输入频率的个数与目标刺激序列也相同时，确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。具体地，用户初始化设置身份识别系统时，用户选取[12,8,20]Hz作为目标刺激序列，并将三个频率刺激下的自己真实脑电信号X(12,t)、X(8,t)与X(20,t)保存到保险箱身份识别系统中。当待检测用户请求身份认证时，系统要求待检测用户输入刺激频率序列指令以#键结束。若待检测用户输入[6,12,15#]等非设定频率大小序列，或输入[12,8#]等个数错误的序列，均说明待检测用户身份验证失败，结束身份识别。当用户正确输入[12,8,20#]时，或者输入
[12,20,8#]等仅仅是顺序错误的序列，都可以说明待检测用户本次关于刺激频率序列输入的身份验证成功，可以继续进行关于SSVEP信号的身份验证。
可选地，待检测用户输入刺激频率序列[f′1,f′2,f′3,……,f′n]，在确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]是否相同时，还可以不考虑输入个数，即在输入的每个频率的大小与目标刺激序列中对应频率大小相同以及输入频率的顺序与目标刺激序列中频率顺序也相同时，或者仅在输入的每个频率的大小与目标刺激序列中对应频率大小相同，均可确定刺激频率序列[f′1,f′2,f′3,……,f′n]与目标刺激序列[f1,f2,f3,……,fn]相同。具体地，用户在设置初始的刺激频率序列时，可以设置大量频率，例如设置10个频率构成初始刺激频率序列。当待检测用户要进行身份验证时，可以只输入少量频率构成的刺激频率序列，若该刺激频率序列包括的频率均属于用户设置的初始的刺激频率序列时，身份识别通过，否则不通过。
例如，用户设置10个频率构成的初始刺激频率序列，当待检测用户进行身份识别时，输入刺激频率序列[f′1,f′2,f′3,……,f′n]为[12,8,20]，若查找初始刺激频率序列，确定包括该三个频率时，即存在目标刺激频率序列[f1,f2,f3,……,fn]为[12,8,20]，该目标刺激序列为用户设置的10个频率集合的子集，该[f′1,f′2,f′3,……,f′n]与[f1,f2,f3,……,fn]相同，则身份识别通过，继续进行关于SSVEP信号的身份识别；若查找初始刺激频率序列，确定不包括该三个频率或只包括其中一个或两个频率时，例如，只存在目标刺激频率序列[f1,f2,f3,……,fn]为[12,8]，不包括频率20Hz，则该[f′1,f′2,f′3,……,f′n]与[f1,f2,f3,……,fn]不同，身份识别不通过，可以结束身份识别。
再例如，在上述实例中，为了减少误差，还可以进一步考虑顺序是否正确。同样的，用户设置10个频率构成的初始刺激频率序列，当待检测用户进行身份识 别时，输入刺激频率序列[f′1,f′2,f′3,……,f′n]为[12,8,20]，若查找初始刺激频率序列，确定包括连续的该三个频率，且该三个频率顺序相同时，即10个频率中存在一段频率序列，即目标刺激频率序列[f1,f2,f3,……,fn]为[12,8,20]，该[f′1,f′2,f′3,……,f′n]与[f1,f2,f3,……,fn]相同，则身份识别通过，继续进行关于SSVEP信号的身份识别；若查找初始刺激频率序列，确定不包括连续的该三个频率，或只包括其中一个或两个频率，或顺序不同时，例如，只存在目标刺激频率序列[f1,f2,f3,……,fn]为[12,8]，不包括频率20Hz，或存在目标刺激频率序列为[12,20,8]，即输入顺序不同，或存在目标刺激频率序列为[12,11,20,8]，即输入顺序不连续，则都说明该[f′1,f′2,f′3,……,f′n]与[f1,f2,f3,……,fn]不同，身份识别不通过，结束身份识别。
在本发明实施例中，该目标刺激频率序列[f1,f2,f3,……,fn]也可以不作为身份识别依据，当待检测用户需要进行身份识别时，由身份识别系统自动获取事先由用户设置的目标刺激频率序列[f1,f2,f3,……,fn]，或者随机获取或按顺序获取事先由用户设置的一系列频率中部分频率，构成目标刺激频率序列[f1,f2,f3,……,fn]，通过获取到的该目标刺激频率序列[f1,f2,f3,……,fn]对待检测用于进行SSVEP信号的身份识别。
S120，为待检测用户显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，该n段刺激信号中第i段刺激信号的显示频率为该目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n。
S130，获取所述待检测用户由于所述n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号。
在本发明实施例中，待检测用户进行身份识别时，确定目标刺激频率序列[f1,f2,f3,……,fn]后，根据该目标刺激频率序列[f1,f2,f3,……,fn]为待检测用户依次显示n段对应的刺激信号，该n段刺激信号中第i段刺激信号的显示频率为该目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i依次取1、2、3……n。具体地，第一段刺激信号以f1为频率显示一段时间，然后第二段刺激信号以f2为频率显示一段时间，第三段刺激信号以f3为频率显示一段时间，一直到第n段刺激信号以fn为频率显示一段时间，则可以采集到待检测用户根据该n段刺激信号产生的对应的n段SSVEP信号。
应理解，类似的，用户在设置预设SSVEP信号时，也是根据依次显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，产生n段SSVEP信号，采集该n段SSVEP信号作为预设SSVEP信号进行保存，以便于根据预设SSVEP信号检测待检测用户的SSVEP信号是否正确，从而进行身份验证。
可选地，为待检测用户或设置SSVEP信号的原始用户显示的刺激信号的显示频率为fi，这里的刺激信号可以为带有一定图案的图片，该图片按照对应的频率fi显示；或者该刺激信号也可以为灯光，灯光的开关或明暗交替的频率即为显示频率fi；或者，该刺激信号还可以为其他任何可以诱发人脑产生SSVEP信号的形式， 通过频率fi向用户显示，以便于获取该待检测用户或原始用户对于刺激信号的反应产生的SSVEP信号。
应理解，设置预设SSVEP信号的原始用户使用的刺激信号类型与待检测用户使用的刺激信号的类型一致，这样可以避免由于刺激信号不同带来的误差。
应理解，采集设置预设SSVEP信号的原始用户或待检测用户的SSVEP信号时，可以让用户带上可穿戴干电极帽，该电极帽至少包括一个置于用户枕区头皮处的干电极，以便于采集用户的SSVEP信号。在采集待检测用户或原始用户根据刺激信号产生的SSVEP信号时，用户用眼睛注视刺激信号并保持注意力。此时，用户大脑枕区诱发的SSVEP信号便可通过可穿戴脑电设备(即可穿戴干电极帽)采集到。采集到的SSVEP信号可被无线传输到相应的信号分析区域，供后续分析处理识别。
在本发明实施例中，在每段刺激信号开始时，可以向采集SSVEP信号的相应模块发送同步信息，以便于在采集SSVEP信号时，可以根据该同步信号截取不同频率对应的每段刺激信号的SSVEP信号，即n段刺激信号对应n段SSVEP信号，通过该同步信号可以分别截取出该n段SSVEP信号。
S140，当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定该待检测用户身份正确；当该n段SSVEP信号与该n段预设SSVEP信号的相似度小于该阈值时，确定该待检测用户身份错误。
在本发明实施例中，对于采集到的n段SSVEP信号，每段SSVEP信号对应一个刺激信号的频率，即第i段SSVEP信号可以表示为Xi(fi,Ti)，fi表示该第i段SSVEP信号是根据显示频率为fi的刺激信号获得的，即目标刺激频率序列[f1,f2,f3,……,fn]对应n段待检测用户的SSVEP信号X1(f1,T1)、X2(f2,T2)……Xn(fn,Tn)。相应地，用户事先设置的预设SSVEP信号可以表示为X′i(fi,Ti)，即目标刺激频率序列[f1,f2,f3,……,fn]对应n段预设SSVEP信号X′1(f1,T1)、X′2(f2,T2)……X′n(fn,Tn)。
应理解，当采集到的SSVEP信号是多个电极的信号时，Xi(fi,Ti)或者X′i(fi,Ti)可以为一个矩阵。此时，可以用主成分分析(PCA)或典型相关分析(CCA)等经典数据降维方法，将Xi(fi,Ti)或者X′i(fi,Ti)降为一维时域信号。以下将Xi(fi,Ti)和X′i(fi,Ti)视为一维时域信号为例进行处理。
在本发明实施例中，将采集到的待检测用户的SSVEP信号与预设SSVEP信号进行对比分析，进而确定待检测用户身份是否正确。具体地，可以通过计算该待检测用户的n段SSVEP信号与n段预设SSVEP信号的相似度，确定身份识别结果。当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，说明待检测用户的身份正确，则可以通过身份识别；当该n段SSVEP信号与n段预设SSVEP信号的相似度小于阈值时，说明该待检测用户的身份错误，身份识别不通过。例如，通过该身份识别开启保险柜，则身份识别通过，即可以开启保险 柜；但身份识别不通过时，则不可以开启保险柜，并且进一步的，在身份识别不通过时，还可以启动报警装置。
应理解，可以通过n段SSVEP信号与n段SSVEP信号的相关系数，计算二者的相似度。可选地，可以通过时域角度计算n段SSVEP信号与n段预设SSVEP信号的相关系数。具体地，可以计算第i段SSVEP信号Xi(fi,Ti)与第i段预设SSVEP信号X′i(fi,Ti)的时域相关系数rX(fi)，该相关系数可以为线性相关系数，例如，皮尔森线性相关系数。皮尔森线性相关系数，可以用来衡量输入信号与真实用户信号之间的相似程度。即使在相同的刺激频率下，不同人由于大脑生理结构的不同，例如，视觉通路延迟不同，视觉系统的频谱响应不同等，其各自诱发的SSVEP信号之间会存在着较大差异。因此，只有当输入的SSVEP信号是真实用户本人的脑电信号时，相关系数才会较大。
可选地，还可以通过变换域角度计算n段SSVEP信号与n段预设SSVEP信号的变换域相关系数。具体地，可以将n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；将n段预设SSVEP信号中的X′i(fi,Ti)进行变换得到Z′i(fi,Yi)；计算确定Zi(fi,Yi)与Z′i(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；根据所述变换域相关系数rY(fi)，确定所述相关系数向量
Figure PCTCN2016104445-appb-000023
应理解，变换域是指不同于时域(T域)的其它域，例如，频域(即F域)、S域、Z域等。可以通过算法将时域的SSVEP信号Xi(fi,Ti)变换到变换域，得到Zi(fi,Yi)；例如，可以通过傅里叶算法将SSVEP信号Xi(fi,Ti)变换到频域，得到Zi(fi,Fi)，还可以通过快速傅里叶算法将SSVEP信号Xi(fi,t)变换到频域，得到又一种Zi(fi,Fi)；再例如，通过拉普拉斯算法将SSVEP信号Xi(fi,t)变换到S域，得到Zi(fi,Si)。
例如，该变换域相关系数rY(fi)可以包括频域相关系数rF(fi)，即可以通过频域角度计算n段SSVEP信号与n段预设SSVEP信号的频域相关系数。例如，以计算幅频响应的相似性为例。具体地，将第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换，得到其对应的幅频响应信号Zi(fi,Fi)，同样的，将第i段预设SSVEP信号X′i(fi,Ti)也进行傅里叶变换，得到Z′i(fi,Fi)，计算Zi(fi,Fi)与Z′i(fi,Fi)的频域相关系数为rF(fi)。类似的，还可以计算n段SSVEP信号与n段预设SSVEP信号的相频响应的相似系数，本发明实施例并不限于此。
在本发明实施例中，可以根据上述两种方法确定的时域相关系数rX(fi)以及变换域相关系数为rY(fi)，确定该n段SSVEP信号与n段预设SSVEP信号的相似度。具体地，可以根据n个时域相关系数rX(fi)以及n个变换域相关系数为rY(fi)，确定相关系数向量
Figure PCTCN2016104445-appb-000024
该相关系数向量
Figure PCTCN2016104445-appb-000025
中的元素可以表示时域相关系数，也可以表示变换域相关系数，其中，该变换域相关系数可以为频域相关系数，或者还可以为其他相关系数。例如，该相关系数向量
Figure PCTCN2016104445-appb-000026
可以包括n个元素，该n个元素可以为n个时域相关系数rX(fi)，也可以为n个变换域相关系数为rY(fi)，或者，该n个 元素可以包括i个时域相关系数rX(fi)，以及包括剩余(n-i)个时域相关系数rX(fi)对应的(n-i)个变换域相关系数为rY(fi)；再例如，该相关系数向量
Figure PCTCN2016104445-appb-000027
也可以包括2n个元素，该2n个元素包括n个时域相关系数rX(fi)以及n个变换域相关系数为rY(fi)，本发明并不限于此。
在本发明实施例中，根据该相关系数向量
Figure PCTCN2016104445-appb-000028
确定该n段SSVEP信号与n段预设SSVEP信号的相似度。具体地，可以通过公式
Figure PCTCN2016104445-appb-000029
确定n段SSVEP信号与n段预设SSVEP信号的相似度，其中，
Figure PCTCN2016104445-appb-000030
表示权重参数向量，即该n段SSVEP信号与n段预设SSVEP信号的相似度等于权重参数向量
Figure PCTCN2016104445-appb-000031
的转置与相关系数向量
Figure PCTCN2016104445-appb-000032
的内积。
具体地，该权重参数向量
Figure PCTCN2016104445-appb-000033
中每个元素，对应于相关系数向量
Figure PCTCN2016104445-appb-000034
中每个元素表示的相关系数的权重。例如，该权重参数向量
Figure PCTCN2016104445-appb-000035
中每个元素均设置为等于
Figure PCTCN2016104445-appb-000036
其中，
Figure PCTCN2016104445-appb-000037
表示相关系数向量
Figure PCTCN2016104445-appb-000038
中包括的元素的个数。或者，当统计显示该相关系数向量
Figure PCTCN2016104445-appb-000039
中某个元素，相比于其他元素可以更好的反映真实用户与非真实用户的差异时，则可以将该元素对应的权重值调大。
应理解，这里的该元素可以更好的反映真实用户与非真实用户的差异，指的是非真实用户与原始真实用户的相关系数进行比较时，该元素的差值相比于其他元素较大。具体地，原始用户在设置预设SSVEP信号时，可以进行多次采集，确定预设SSVEP信号后，每次采集到的该原始用户的SSVEP信号与预设的SSVEP信号都可以确定一个相关系数向量
Figure PCTCN2016104445-appb-000040
进行多次采集，即可以确定多组相关系数向量
Figure PCTCN2016104445-appb-000041
多组相关系数向量
Figure PCTCN2016104445-appb-000042
的平均值可作为原始用户的原始相关系数向量
Figure PCTCN2016104445-appb-000043
可以反映出原始用户本身与每段预设SSVEP信号的相似性，即。相应地，在待检测用户进行身份识别时，确定待检测用户的SSVEP信号与预设SSVEP信号之间的相关系数向量
Figure PCTCN2016104445-appb-000044
当统计显示非真实用户该相关系数向量
Figure PCTCN2016104445-appb-000045
中某个元素和原始相关系数向量
Figure PCTCN2016104445-appb-000046
对应的元素的差值，大于其他元素和原始相关系数向量
Figure PCTCN2016104445-appb-000047
对应的元素之间的差值，说明该元素可以较强的反映了的原始用户与待检测用户的区分性，因此可以调制该元素对应的权重值偏大。
例如，待检测用户进行身份识别时，目标刺激频率序列为[12,8]，则对应的计算时域相关系数r(12)和r(8)，令相关系数向量
Figure PCTCN2016104445-appb-000048
中只包括该两个相关系数，即r(12)和r(8)，则可以设置
Figure PCTCN2016104445-appb-000049
可选地，若统计数据显示计算的非真实用户的r(12)和r(8)中，r(12)与真实原始用户的相关系数值R(12)差异更大，则可以修改权重值，令
Figure PCTCN2016104445-appb-000050
其中，确定该r(12)差异较大是指，例如，原始用户在设置预设SSVEP信号时，确定的R(12)和R(8)均为0.9，但是计算的待检测用户r(8)为0.5，r(12)为0.2，r(12)与R(12)的差值0.7大于r(8)与R(8)的差值0.4，则可以加大该r(12)对应的权重，即令
Figure PCTCN2016104445-appb-000051
但是，若同样的，待检测用户的r(12)为0.2，但是原始用户在设置预设SSVEP信号时，确定的R(12)为0.8，而待检测用户的r(8)为 0.3，原始用户在设置预设SSVEP信号时，确定的R(8)为0.9，此时，待检测用户的r(12)与R(12)相差0.6，r(8)与R(8)相差也是0.6，因此可以不加大该待检测用户的r(12)的权重值，令
Figure PCTCN2016104445-appb-000052
在本发明实施例中，根据该待检测用户的n段SSVEP信号与n段预设SSVEP信号的相似度，进行身份识别。当该相似度大于或等于阈值时，说明该待检测用户身份正确，即通过本次基于SSVEP信号的身份识别；当该相似度小于阈值时，说明该待检测用户身份错误，即不能通过本次基于SSVEP信号的身份识别。可选地，该阈值可以根据实际情况进行设置，并且可以根据多次测试，不断更新该阈值。
应理解，在本发明的各种实施例中，上述各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。
因此，本发明实施例的基于脑电信号的身份识别的方法，通过设置目标刺激频率序列，按照一定频率为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。另外，对于设置的目标刺激频率序列，也可以进行身份识别，这样，专属的刺激序列加SSVEP信号双认证的模式，提高了设备保密的等级，更不易被侵入。
上文中结合图1，详细描述了根据本发明实施例的基于脑电信号的身份识别的方法，下面将结合图2至图3，描述根据本发明实施例的基于脑电信号的身份识别的装置。
如图2所示，根据本发明实施例的基于脑电信号的身份识别的装置200包括：
确定单元210，用于确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数；
显示单元220，用于为待检测用户显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，该n段刺激信号中第i段刺激信号的显示频率为该目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n；
获取单元230，用于获取该待检测用户由于该n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号；
处理单元240，用于当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定该待检测用户身份正确；当该n段SSVEP信号与该n段预 设SSVEP信号的相似度小于该阈值时，确定该待检测用户身份错误。
因此，本发明实施例的基于脑电信号的身份识别的装置，根据目标刺激频率序列为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。
可选地，该确定单元210具体用于：获取该待检测用户输入的待检测刺激频率序列[f′1,f′2,f′3,……,f′n]；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]相同时，获取该目标刺激频率序列[f1,f2,f3,……,fn]对应的该n段刺激信号；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]不同时，确定该待检测用户身份错误。
可选地，该处理单元240具体用于：根据至少一个域的该n段SSVEP信号与对应的至少一个域的该n段预设SSVEP信号，确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000053
该相关系数向量
Figure PCTCN2016104445-appb-000054
中的每个元素表示该n段SSVEP信号与该n段预设SSVEP信号的相关系数；将
Figure PCTCN2016104445-appb-000055
确定为该n段SSVEP信号与该n段预设SSVEP信号的相似度，其中，
Figure PCTCN2016104445-appb-000056
表示权重参数向量，该权重参数向量
Figure PCTCN2016104445-appb-000057
中的每个元素表示该相关系数向量
Figure PCTCN2016104445-appb-000058
中对应元素的权重值。
可选地，该处理单元240具体用于：确定该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)与该n段预设SSVEP信号中的X′i(fi,Ti)的时域相关系数为rX(fi)，i取1、2、3……n；根据该时域相关系数rX(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000059
可选地，该处理单元240具体用于：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；将该n段预设SSVEP信号中的X′i(fi,Ti)进行变换得到Z′i(fi,Yi)；确定Zi(fi,Yi)与Z′i(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；根据该变换域相关系数rY(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000060
可选地，该处理单元240具体用于：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换得到Zi(fi,Fi)；将该n段预设SSVEP信号中的X′i(fi,Ti)进行傅里叶变换得到Z′i(fi,Fi)；确定Zi(fi,Fi)与Z′i(fi,Fi)的频域相关系数为rF(fi)，i取1、2、3……n；根据该频域相关系数rF(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000061
应理解，根据本发明实施例的基于脑电信号的身份识别的装置200可对应于执行本发明实施例中的方法100，并且装置200中的各个模块的上述和其它操作和/或功能分别为了实现图1中的各个方法的相应流程，为了简洁，在此不再赘述。
因此，本发明实施例的基于脑电信号的身份识别的装置，根据目标刺激频率序列为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。另外，对于设置的目标刺激频率序列，也可以进行身份识别，这样，专属的刺激序列加SSVEP信号双认证的模式，提高了设备保密的等级，更不易被侵入。
如图3所示，本发明实施例还提供了一种基于脑电信号的身份识别的装置300，包括处理器310和存储器320，还可以包括总线系统330。其中，处理器310和存储器320通过总线系统330相连，该存储器320用于存储指令，该处理器310用于执行该存储器320存储的指令。该存储器320存储程序代码，且处理器310可以调用存储器320中存储的程序代码执行以下操作：确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数；为待检测用户显示该目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，该n段刺激信号中第i段刺激信号的显示频率为该目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n；获取该待检测用户由于该n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号；当该n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定该待检测用户身份正确；当该n段SSVEP信号与该n段预设SSVEP信号的相似度小于该阈值时，确定该待检测用户身份错误。
因此，本发明实施例的基于脑电信号的身份识别的装置，根据目标刺激频率序列为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。
应理解，在本发明实施例中，该处理器310可以是中央处理单元(Central Processing Unit，简称为“CPU”)，该处理器310还可以是其他通用处理器、数字 信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
该存储器320可以包括只读存储器和随机存取存储器，并向处理器310提供指令和数据。存储器320的一部分还可以包括非易失性随机存取存储器。例如，存储器320还可以存储设备类型的信息。
该总线系统330除包括数据总线之外，还可以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起见，在图中将各种总线都标为总线系统330。
在实现过程中，上述方法的各步骤可以通过处理器310中的硬件的集成逻辑电路或者软件形式的指令完成。结合本发明实施例所公开的方法的步骤可以直接体现为硬件处理器执行完成，或者用处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器320，处理器310读取存储器320中的信息，结合其硬件完成上述方法的步骤。为避免重复，这里不再详细描述。
可选地，该处理器310用于：获取该待检测用户输入的待检测刺激频率序列[f′1,f′2,f′3,……,f′n]；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]相同时，获取该目标刺激频率序列[f1,f2,f3,……,fn]对应的该n段刺激信号；当该待检测刺激频率序列[f′1,f′2,f′3,……,f′n]与该目标刺激频率序列[f1,f2,f3,……,fn]不同时，确定该待检测用户身份错误。
可选地，该处理器310用于：根据至少一个域的该n段SSVEP信号与对应的至少一个域的该n段预设SSVEP信号，确定该n段SSVEP信号与该n段预设SSVEP信号的相关系数向量
Figure PCTCN2016104445-appb-000062
该相关系数向量
Figure PCTCN2016104445-appb-000063
中的每个元素表示该n段SSVEP信号与该n段预设SSVEP信号的相关系数；将
Figure PCTCN2016104445-appb-000064
确定为该n段SSVEP信号与该n段预设SSVEP信号的相似度，其中，
Figure PCTCN2016104445-appb-000065
表示权重参数向量，该权重参数向量
Figure PCTCN2016104445-appb-000066
中的每个元素表示该相关系数向量
Figure PCTCN2016104445-appb-000067
中对应元素的权重值。
可选地，该处理器310用于：确定该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)与该n段预设SSVEP信号中的X′i(fi,Ti)的时域相关系数为rX(fi)，i取1、2、3……n；根据该时域相关系数rX(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000068
可选地，该处理器310用于：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；将该n段预设SSVEP信号中的X′i(fi,Ti)进行变换得到Z′i(fi,Yi)；确定Zi(fi,Yi)与Z′i(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；根据该变换域相关系数rY(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000069
可选地，该处理器310用于：将该n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换得到Zi(fi,Fi)；将该n段预设SSVEP信号中的X′i(fi,Ti)进行傅里叶变换得到Z′i(fi,Fi)；确定Zi(fi,Fi)与Z′i(fi,Fi)的频域相关系数为rF(fi)，i 取1、2、3……n；根据该频域相关系数rF(fi)，确定该相关系数向量
Figure PCTCN2016104445-appb-000070
应理解，根据本发明实施例的基于脑电信号的身份识别的装置300可对应于本发明实施例中的基于脑电信号的身份识别的装置200，并可以对应于执行根据本发明实施例的方法100中的相应主体，并且装置300中的各个模块的上述和其它操作和/或功能分别为了实现图1中的各个方法的相应流程，为了简洁，在此不再赘述。
因此，本发明实施例的基于脑电信号的身份识别的装置，根据目标刺激频率序列为待检测用户显示刺激信号，从而采集待检测用户产生的SSVEP信号，将该SSVEP信号与预设SSVEP信号进行对比，进行待检测用户的身份识别。由于SSVEP信号高信噪比的特点可以使得信号更容易检测，相比于现有脑电身份识别方法可以缩短刺激时长；SSVEP信号主要集中在人的大脑枕区，仅需要较少的电极，比如一个电极就可以采集到丰富信息量的信号，使用方便；SSVEP信号作为一种初级视觉皮层诱发的信号，不需要人的高级认知活动参与，因此其受人的精神状态影响较小，信号特征相对更稳定；SSVEP信号丰富的幅度频率响应以及相位频率响应可以直接与人初级视觉皮层的系统生理特性对应，使得基于此构建的身份识别系统更加的保密，不易被复制伪造。另外，对于设置的目标刺激频率序列，也可以进行身份识别，这样，专属的刺激序列加SSVEP信号双认证的模式，提高了设备保密的等级，更不易被侵入。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来 实现本实施例方案的目的。
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应所述以权利要求的保护范围为准。

Claims (12)

  1. 一种基于脑电信号的身份识别的方法，其特征在于，包括：
    确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数；
    为待检测用户显示所述目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，所述n段刺激信号中第i段刺激信号的显示频率为所述目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n；
    获取所述待检测用户由于所述n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号；
    当所述n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定所述待检测用户身份正确；
    当所述n段SSVEP信号与所述n段预设SSVEP信号的相似度小于所述阈值时，确定所述待检测用户身份错误。
  2. 根据权利要求1所述的方法，其特征在于，所述确定目标刺激频率序列[f1,f2,f3,……,fn]，包括：
    获取所述待检测用户输入的待检测刺激频率序列[f1′,f2′,f3′,……,fn′]；
    当所述待检测刺激频率序列[f1′,f2′,f3′,……,fn′]与所述目标刺激频率序列[f1,f2,f3,……,fn]相同时，获取所述目标刺激频率序列[f1,f2,f3,……,fn]对应的所述n段刺激信号；
    当所述待检测刺激频率序列[f1′,f2′,f3′,……,fn′]与所述目标刺激频率序列[f1,f2,f3,……,fn]不同时，确定所述待检测用户身份错误。
  3. 根据权利要求1或2所述的方法，其特征在于，所述方法还包括：
    根据至少一个域的所述n段SSVEP信号与对应的至少一个域的所述n段预设SSVEP信号，确定所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数向量
    Figure PCTCN2016104445-appb-100001
    所述相关系数向量
    Figure PCTCN2016104445-appb-100002
    中的每个元素表示所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数；
    将
    Figure PCTCN2016104445-appb-100003
    确定为所述n段SSVEP信号与所述n段预设SSVEP信号的相似度，其中，
    Figure PCTCN2016104445-appb-100004
    表示权重参数向量，所述权重参数向量
    Figure PCTCN2016104445-appb-100005
    中的每个元素表示所述相关系数向量
    Figure PCTCN2016104445-appb-100006
    中对应元素的权重值。
  4. 根据权利要求3所述的方法，其特征在于，所述确定所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数向量
    Figure PCTCN2016104445-appb-100007
    包括：
    确定所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)与所述n段预设SSVEP信号中的Xi′(fi,Ti)的时域相关系数为rX(fi)，i取1、2、3……n；
    根据所述时域相关系数rX(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100008
  5. 根据权利要求3或4所述的方法，其特征在于，所述确定所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数向量
    Figure PCTCN2016104445-appb-100009
    包括：
    将所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；
    将所述n段预设SSVEP信号中的Xi′(fi,Ti)进行变换得到Zi′(fi,Yi)；
    确定Zi(fi,Yi)与Zi′(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；
    根据所述变换域相关系数rY(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100010
  6. 根据权利要求3或4所述的方法，其特征在于，所述确定所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数向量
    Figure PCTCN2016104445-appb-100011
    包括：
    将所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换得到Zi(fi,Fi)；
    将所述n段预设SSVEP信号中的Xi′(fi,Ti)进行傅里叶变换得到Zi′(fi,Fi)；
    确定Zi(fi,Fi)与Zi′(fi,Fi)的频域相关系数为rF(fi)，i取1、2、3……n；
    根据所述频域相关系数rF(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100012
  7. 一种基于脑电信号的身份识别的装置，其特征在于，包括：
    确定单元，用于确定目标刺激频率序列[f1,f2,f3,……,fn]，n为正整数；
    显示单元，用于为待检测用户显示所述目标刺激频率序列[f1,f2,f3,……,fn]对应的n段刺激信号，所述n段刺激信号中第i段刺激信号的显示频率为所述目标刺激频率序列[f1,f2,f3,……,fn]中第i个频率fi，i取1、2、3……n；
    获取单元，用于获取所述待检测用户由于所述n段刺激信号产生的n段稳态视觉诱发电位SSVEP信号；
    处理单元，用于当所述n段SSVEP信号与n段预设SSVEP信号的相似度大于或等于阈值时，确定所述待检测用户身份正确；当所述n段SSVEP信号与所述n段预设SSVEP信号的相似度小于所述阈值时，确定所述待检测用户身份错误。
  8. 根据权利要求7所述的装置，其特征在于，所述确定单元具体用于：
    获取所述待检测用户输入的待检测刺激频率序列[f1′,f2′,f3′,……,fn′]；
    当所述待检测刺激频率序列[f1′,f2′,f3′,……,fn′]与所述目标刺激频率序列[f1,f2,f3,……,fn]相同时，获取所述目标刺激频率序列[f1,f2,f3,……,fn]对应的所述n段刺激信号；
    当所述待检测刺激频率序列[f1′,f2′,f3′,……,fn′]与所述目标刺激频率序列[f1,f2,f3,……,fn]不同时，确定所述待检测用户身份错误。
  9. 根据权利要求7或8所述的装置，其特征在于，所述处理单元具体用于：
    根据至少一个域的所述n段SSVEP信号与对应的至少一个域的所述n段预设SSVEP信号，确定所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数向量
    Figure PCTCN2016104445-appb-100013
    所述相关系数向量
    Figure PCTCN2016104445-appb-100014
    中的每个元素表示所述n段SSVEP信号与所述n段预设SSVEP信号的相关系数；
    将
    Figure PCTCN2016104445-appb-100015
    确定为所述n段SSVEP信号与所述n段预设SSVEP信号的相似度，其中，
    Figure PCTCN2016104445-appb-100016
    表示权重参数向量，所述权重参数向量
    Figure PCTCN2016104445-appb-100017
    中的每个元素表示所述相关系数向量
    Figure PCTCN2016104445-appb-100018
    中对应元素的权重值。
  10. 根据权利要求9所述的装置，其特征在于，所述处理单元具体用于：
    确定所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)与所述n段预设SSVEP信号中的Xi′(fi,Ti)的时域相关系数为rX(fi)，i取1、2、3……n；
    根据所述时域相关系数rX(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100019
  11. 根据权利要求9或10所述的装置，其特征在于，所述处理单元具体用于：
    将所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行变换得到Zi(fi,Yi)；
    将所述n段预设SSVEP信号中的Xi′(fi,Ti)进行变换得到Zi′(fi,Yi)；
    确定Zi(fi,Yi)与Zi′(fi,Yi)的变换域相关系数为rY(fi)，i取1、2、3……n；
    根据所述变换域相关系数rY(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100020
  12. 根据权利要求9或10所述的装置，其特征在于，所述处理单元具体用于：
    将所述n段SSVEP信号中第i段SSVEP信号Xi(fi,Ti)进行傅里叶变换得到Zi(fi,Fi)；
    将所述n段预设SSVEP信号中的Xi′(fi,Ti)进行傅里叶变换得到Zi′(fi,Fi)；
    确定Zi(fi,Fi)与Zi′(fi,Fi)的频域相关系数为rF(fi)，i取1、2、3……n；
    根据所述频域相关系数rF(fi)，确定所述相关系数向量
    Figure PCTCN2016104445-appb-100021
PCT/CN2016/104445 2016-05-26 2016-11-03 基于脑电信号的身份识别的方法和装置 WO2017201972A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610361544.X 2016-05-26
CN201610361544.XA CN107437011A (zh) 2016-05-26 2016-05-26 基于脑电信号的身份识别的方法和装置

Publications (1)

Publication Number Publication Date
WO2017201972A1 true WO2017201972A1 (zh) 2017-11-30

Family

ID=60411041

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/104445 WO2017201972A1 (zh) 2016-05-26 2016-11-03 基于脑电信号的身份识别的方法和装置

Country Status (2)

Country Link
CN (1) CN107437011A (zh)
WO (1) WO2017201972A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109816406A (zh) * 2019-02-26 2019-05-28 北京理工大学 一种物品标记方法、装置、设备和介质
CN113197551A (zh) * 2021-05-07 2021-08-03 中国医学科学院生物医学工程研究所 多模生理神经信号检测与实验刺激时间对齐方法
CN114246594A (zh) * 2021-12-17 2022-03-29 天津大学 脑电信号处理方法、背景脑电预测模型的训练方法及装置

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108153414B (zh) * 2017-12-19 2021-06-15 同济大学 稳态视觉诱发电位信号的识别方法、系统及装置
CN108985029A (zh) * 2018-06-05 2018-12-11 中国科学院半导体研究所 基于视觉刺激的脑电身份识别方法
CN109101807B (zh) * 2018-09-10 2022-12-02 清华大学 一种脑电身份权限控制系统及方法
CN109766751B (zh) * 2018-11-28 2022-02-01 西安电子科技大学 基于频域编码的稳态视觉诱发脑电身份识别方法及系统
CN109948700B (zh) * 2019-03-19 2020-07-24 北京字节跳动网络技术有限公司 用于生成特征图的方法和装置
CN111329474B (zh) * 2020-03-04 2021-05-28 西安电子科技大学 基于深度学习的脑电身份识别方法、系统及信息更新方法
CN111714120A (zh) * 2020-05-08 2020-09-29 广东食品药品职业学院 能进行视觉定位能力评估的脑机接口系统及其应用
CN113360876A (zh) * 2021-06-30 2021-09-07 杭州回车电子科技有限公司 基于ssvep的身份识别方法、装置、电子装置和存储介质
CN113935376B (zh) * 2021-10-13 2023-03-10 中国科学技术大学 一种基于联合约束典型相关分析的大脑功能子区划分方法
CN114424945B (zh) * 2021-12-08 2024-05-31 中国科学院深圳先进技术研究院 一种基于随机图形图像闪现的脑波生物特征识别系统与方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101828921A (zh) * 2010-06-13 2010-09-15 天津大学 基于视觉诱发脑电vep的身份识别方法
CN102509103A (zh) * 2011-09-29 2012-06-20 天津大学 基于视觉诱发p3脑电的身份识别
CN102755162A (zh) * 2012-06-14 2012-10-31 天津大学 一种基于视听认知事件相关脑电的身份识别方法
US8483816B1 (en) * 2010-02-03 2013-07-09 Hrl Laboratories, Llc Systems, methods, and apparatus for neuro-robotic tracking point selection
CN103729586A (zh) * 2013-12-20 2014-04-16 北京握奇数据系统有限公司 一种基于脑电波信号的密码验证方法及系统
CN105389494A (zh) * 2015-11-04 2016-03-09 上海大学 基于脑电波的生物识别密码验证系统和方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102866775A (zh) * 2012-09-04 2013-01-09 同济大学 基于多模式融合的脑机接口控制系统及方法
CN105389004A (zh) * 2015-10-22 2016-03-09 上海斐讯数据通信技术有限公司 一种脑电波解锁终端屏幕的系统及方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8483816B1 (en) * 2010-02-03 2013-07-09 Hrl Laboratories, Llc Systems, methods, and apparatus for neuro-robotic tracking point selection
CN101828921A (zh) * 2010-06-13 2010-09-15 天津大学 基于视觉诱发脑电vep的身份识别方法
CN102509103A (zh) * 2011-09-29 2012-06-20 天津大学 基于视觉诱发p3脑电的身份识别
CN102755162A (zh) * 2012-06-14 2012-10-31 天津大学 一种基于视听认知事件相关脑电的身份识别方法
CN103729586A (zh) * 2013-12-20 2014-04-16 北京握奇数据系统有限公司 一种基于脑电波信号的密码验证方法及系统
CN105389494A (zh) * 2015-11-04 2016-03-09 上海大学 基于脑电波的生物识别密码验证系统和方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
PHOTHISONOTHAI, M.: "An Investigation of Using SSVEP for EEG-based User Authentication System", PROCEEDINGS OF APSIPA ANNUAL SUMMIT AND CONFERENCE 2015, 19 December 2015 (2015-12-19), pages 923 - 926, XP032870673 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109816406A (zh) * 2019-02-26 2019-05-28 北京理工大学 一种物品标记方法、装置、设备和介质
CN109816406B (zh) * 2019-02-26 2021-01-22 北京理工大学 一种物品标记方法、装置、设备和介质
CN113197551A (zh) * 2021-05-07 2021-08-03 中国医学科学院生物医学工程研究所 多模生理神经信号检测与实验刺激时间对齐方法
CN113197551B (zh) * 2021-05-07 2023-08-04 中国医学科学院生物医学工程研究所 多模生理神经信号检测与实验刺激时间对齐方法
CN114246594A (zh) * 2021-12-17 2022-03-29 天津大学 脑电信号处理方法、背景脑电预测模型的训练方法及装置
CN114246594B (zh) * 2021-12-17 2024-04-09 天津大学 脑电信号处理方法、背景脑电预测模型的训练方法及装置

Also Published As

Publication number Publication date
CN107437011A (zh) 2017-12-05

Similar Documents

Publication Publication Date Title
WO2017201972A1 (zh) 基于脑电信号的身份识别的方法和装置
Gui et al. A survey on brain biometrics
Jayarathne et al. BrainID: Development of an EEG-based biometric authentication system
CA2835460C (en) System and method for enabling continuous or instantaneous identity recognition based on physiological biometric signals
US10154818B2 (en) Biometric authentication method and apparatus
EP3647976B1 (en) Login method and apparatus, and electronic device
Jayarathne et al. Survey of EEG-based biometric authentication
Vhaduri et al. Biometric-based wearable user authentication during sedentary and non-sedentary periods
KR101643354B1 (ko) 뇌파를 이용한 본인 인증 시스템 및 방법
Dan et al. An identification system based on portable EEG acquisition equipment
Gupta et al. Blink to get in: Biometric authentication for mobile devices using eeg signals
WO2022047272A2 (en) Electronic devices with a static artificial intelligence model for contextual situations, including age blocking for vaping and ignition start, using data analysis and operating methods thereof
Liang et al. Identity recognition using biological electroencephalogram sensors
Gui et al. Multichannel EEG-based biometric using improved RBF neural networks
Martinovic et al. Pulse-response: Exploring human body impedance for biometric recognition
WO2018112799A1 (zh) 一种基于视觉诱发电位的身份识别方法和设备
Attallah Multi-tasks biometric system for personal identification
Boubakeur et al. EEG-based person recognition analysis and criticism
Cao et al. Towards nonintrusive and secure mobile two-factor authentication on wearables
KR101281852B1 (ko) 뇌 신호를 이용한 생체 인증 장치 및 방법
Rathi et al. A comparative study of classification methods for designing a pictorial P300-based authentication system
Rathi et al. Brain signatures perspective for high-security authentication
Thomas et al. EEG-based biometrie authentication using self-referential visual stimuli
Yang The use of EEG signals for biometric person recognition
Enamamu Bioelectrical user authentication

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16902956

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16902956

Country of ref document: EP

Kind code of ref document: A1