技术领域technical field
本申请涉及计算机技术领域，具体涉及生物特征识别技术领域，尤其涉及用于生物认证的方法、装置以及生物认证系统。The present application relates to the field of computer technology, specifically to the field of biometric identification technology, and in particular to a method, device and biometric authentication system for biometric authentication.
背景技术Background technique
随着传感器制造技术和模式识别、机器学习技术的进步，生物特征识别技术得到了更加广泛的普及和发展。然而为了防止利用盗取或伪造的生物特征进行身份认证，身份认证系统需要具有活体检测功能，以确认生物特征来自于有生命的真实的个体。With the advancement of sensor manufacturing technology, pattern recognition and machine learning technology, biometric recognition technology has been more widely popularized and developed. However, in order to prevent the use of stolen or forged biometrics for identity authentication, the identity authentication system needs to have a living body detection function to confirm that the biometrics come from a living and real individual.
现有技术中存在多种基于软件或硬件的用于身份认证系统的活体检测方法，其中使用多种生物特征联合的方法具有较好的防伪造能力，尤其是当这些生物特征中包含难以伪造和复制的、来自生物体内的信号(例如心电信号)时。然而，分别伪造或复制各种生物特征再同时使用，仍然可以欺骗身份认证系统，通过身份认证。因此，需要进一步增强身份认证系统的安全性。In the prior art, there are a variety of liveness detection methods for identity authentication systems based on software or hardware. Among them, the method of combining multiple biometric features has better anti-counterfeiting capabilities, especially when these biometric features include difficult-to-forge and When replicating a signal from a living body (such as an electrocardiographic signal). However, forging or copying various biometrics separately and using them at the same time can still deceive the identity authentication system and pass the identity authentication. Therefore, it is necessary to further enhance the security of the identity authentication system.
发明内容Contents of the invention
本申请提供了用于生物认证的方法、装置以及生物认证系统。The present application provides a biometric authentication method, device and biometric authentication system.
一方面，本申请提供了一种生物认证的方法，该方法包括：接收至少两种生物特征信号；从所述至少两种生物特征信号中分别提取同一生理特征；以及处理所述生理特征以判断所述至少两种生物特征信号是否来自同一真实生物体。In one aspect, the present application provides a biometric authentication method, the method comprising: receiving at least two biometric signals; extracting the same physiological feature from the at least two biometric signals; and processing the physiological features to determine Whether the at least two biometric signals come from the same real organism.
在一些可选的实现方式中，基于来自同一生物体同时采集的至少两种生物特征信号的同一生理特征之间的一致性与来自不同生物体或不同时间采集的至少两种生物特征信号的同一生理特征之间的一致性具有能够区分的差异，进行所述判断。In some optional implementations, based on the consistency between the same physiological characteristics of at least two biological characteristic signals collected simultaneously from the same organism and the identity of at least two biological characteristic signals collected at different times from different organisms The concordance between physiological characteristics has distinguishable differences, making the determination.
在进一步的实现方式中，所述判断包括：基于所述生理特征的属性来计算所述至少两种生物特征信号的生理特征之间的一致性度量；以及响应于所述一致性度量满足预设条件，确认所述至少两种生物特征信号来自同一真实生物体。In a further implementation, the judging includes: calculating a consistency measure between the physiological characteristics of the at least two biometric signals based on the attributes of the physiological characteristics; and responding to the consistency measure satisfying a preset condition, confirming that the at least two biometric signals are from the same authentic organism.
在一些实现方式中，计算一致性度量包括：根据所述生理特征的波形中预定生理现象的采集时间的对应关系来计算一致性度量。In some implementation manners, calculating the consistency metric includes: calculating the consistency metric according to a correspondence between acquisition times of predetermined physiological phenomena in the waveform of the physiological feature.
在一些可选的实现方式中，上述预定生理现象对应于所述生理特征的波形中的波峰或波谷，并且所述一致性度量用差异度或相似度来表征，其中所述差异度表示为所述生理特征的波形中对应波峰或波谷的采集时间偏差的方差，所述相似度表示为所述差异度的倒数。In some optional implementation manners, the aforementioned predetermined physiological phenomenon corresponds to a peak or a trough in the waveform of the physiological feature, and the consistency measure is characterized by a degree of difference or a degree of similarity, wherein the degree of difference is expressed as The variance of the acquisition time deviation corresponding to the peak or trough in the waveform of the physiological feature, the similarity is expressed as the reciprocal of the difference.
在另一些实现方式中，计算一致性度量包括：利用回归器计算所述一致性度量，其中所述回归器经由输入的生理特征数据和设置的一致性度量而训练得到。In some other implementation manners, calculating the consistency measure includes: using a regressor to calculate the consistency measure, wherein the regressor is trained through the input physiological characteristic data and the set consistency measure.
在另一些可选的实现方式中，所述判断包括：基于所述生理特征的属性，利用分类器对所述至少两种生物特征信号进行分类，其中所述分类器使用两类样本训练得到，第一类样本来自同时采集的同一生物体的生物特征信号，第二类样本来自不同时采集的生物特征信号或者来自不同生物体的生物特征信号；以及根据分类结果确认所述至少两种生物特征信号是否来自同一真实生物体。In some other optional implementation manners, the judging includes: using a classifier to classify the at least two types of biological feature signals based on the attributes of the physiological feature, wherein the classifier is trained using two types of samples, The first type of samples come from biometric signals of the same organism collected at the same time, and the second type of samples come from biometric signals collected at different times or from different organisms; and confirming the at least two biometrics according to the classification result Whether the signal is from the same real organism.
在进一步的实现方式中，所述生理特征的属性包括以下至少一项：时域属性，频域属性和统计属性。In a further implementation manner, the attributes of the physiological characteristics include at least one of the following: time domain attributes, frequency domain attributes and statistical attributes.
在进一步的实现方式中，所述时域属性包括所述生理特征中预定生理现象的发生时刻、变化时刻、延续时间或所述生理特征的信号波形；所述频域属性包括所述生理特征的信号频率或频谱分布。In a further implementation manner, the time-domain attribute includes the occurrence moment, change moment, and duration of the predetermined physiological phenomenon in the physiological feature, or the signal waveform of the physiological feature; the frequency-domain attribute includes the signal waveform of the physiological feature Signal frequency or spectral distribution.
在进一步的实现方式中，所述生理特征为随时间变化的生理特征。在进一步的实现方式中，所述生理特征包括心跳和/或呼吸。In a further implementation manner, the physiological characteristics are physiological characteristics that change over time. In a further implementation, the physiological characteristics include heartbeat and/or respiration.
在一些可选的实现方式中，所述用于生物认证的方法还包括：基于所述至少两种生物特征信号是否来自同一真实生物体的判断结果，进行身份认证或识别。In some optional implementation manners, the method for biometric authentication further includes: performing identity authentication or identification based on a judgment result of whether the at least two biometric signals come from the same real biological body.
在进一步的实现方式中，所述基于所述至少两种生物特征信号是否来自同一真实生物体的判断结果，进行身份认证或识别，包括：将从所述至少两种生物特征信号中提取出的身份特征信息与已注册的身份特征信息进行匹配；响应于匹配成功并且所述判断结果确认所述至少两种生物特征信号来自同一真实生物体，认证或识别所述生物体的身份。In a further implementation manner, performing identity authentication or identification based on the judgment result of whether the at least two biometric signals come from the same real biological body includes: extracting the at least two biometric signal The identity feature information is matched with the registered identity feature information; in response to the matching being successful and the judgment result confirms that the at least two biometric signals come from the same real biological body, the identity of the biological body is authenticated or identified.
在进一步的实现方式中，所述身份特征信息包括以下至少一项：人脸图像、指纹图像、掌纹图像、血管图像、虹膜图像、视网膜图像、语音信号、步态特征、签字或笔迹特征、心电信号和脑电信号。In a further implementation, the identity feature information includes at least one of the following: face images, fingerprint images, palmprint images, blood vessel images, iris images, retinal images, voice signals, gait features, signature or handwriting features, ECG and EEG signals.
在一些可选的实现方式中，所述至少两种生物特征信号是同时采集的。In some optional implementation manners, the at least two biometric signals are collected simultaneously.
在进一步的实现方式中，所述采集持续一预定时间段。In a further implementation, the acquisition lasts for a predetermined period of time.
在一些可选的实现方式中，所述生物特征信号包括以下中的至少一项：人脸图像、指纹图像、掌纹图像、血管图像、虹膜图像、视网膜图像、心电信号、脑电信号、光电容积脉搏波(PPG)信号、血压信号、心音信号、人体调制的电磁波信号、胸或腹运动信号和人体导电性信号。In some optional implementation manners, the biometric signal includes at least one of the following: face image, fingerprint image, palmprint image, blood vessel image, iris image, retinal image, ECG signal, EEG signal, Photoplethysmography (PPG) signal, blood pressure signal, heart sound signal, electromagnetic wave signal modulated by the human body, chest or abdominal movement signal and human conductivity signal.
第二方面，本申请提供了一种用于生物认证的装置，所述装置包括：接收单元，配置用于接收至少两种生物特征信号；提取单元，配置用于从所述至少两种生物特征信号中分别提取同一生理特征；以及判断单元，配置用于处理所述生理特征以判断所述至少两种生物特征信号是否来自同一真实生物体。该装置还可以包括配置用于执行根据本申请第一方面所述的方法的各实施方式的步骤的单元或装置。In a second aspect, the present application provides a device for biometric authentication, the device comprising: a receiving unit configured to receive at least two biometric signals; an extraction unit configured to obtain the at least two biometric signals The same physiological feature is extracted from the signals respectively; and the judging unit is configured to process the physiological feature to judge whether the at least two biological feature signals come from the same real organism. The device may further include a unit or device configured to perform the steps of the various implementations of the method according to the first aspect of the present application.
第三方面，本申请提供了一种生物认证系统，包括传感器和处理器，所述传感器配置用于采集至少两种生物特征信号；并且所述处理器配置用于从接收至少两种生物特征信号，从所述至少两种生物特征信号中分别提取同一生理特征，以及处理所述生理特征以判断所述至少两种生物特征信号是否来自同一真实生物体。In a third aspect, the present application provides a biometric authentication system, including a sensor and a processor, the sensor is configured to collect at least two biometric signals; and the processor is configured to receive at least two biometric signals from , respectively extracting the same physiological feature from the at least two biological feature signals, and processing the physiological feature to determine whether the at least two biological feature signals come from the same real organism.
在一些实现方式中，传感器配置用于同时采集所述至少两种生物特征信号。该处理器还可以配置用于执行根据本申请第一方面所述的方法的各实施方式的步骤。In some implementations, the sensor is configured to simultaneously acquire the at least two biometric signals. The processor may also be configured to execute the steps of the various implementations of the method according to the first aspect of the present application.
本申请提供的用于生物认证的方法、装置以及生物认证系统，通过接收至少两种生物特征信号，而后从所述至少两种生物特征信号中分别提取同一生理特征，最后处理所述生理特征以判断所述至少两种生物特征信号是否来自同一真实生物体，实现了待认证或待识别的生物体的真实性的判断，提升了身份认证系统的安全性。The method, device, and biometric authentication system provided by the present application receive at least two biometric signals, then extract the same physiological feature from the at least two biometric signals, and finally process the physiological features to Judging whether the at least two biometric signals come from the same real biological body realizes the judgment of the authenticity of the biological body to be authenticated or identified, and improves the security of the identity authentication system.
附图说明Description of drawings
通过阅读参照以下附图所作的对非限制性实施例所作的详细描述，本申请的其它特征、目的和优点将会变得更明显：Other characteristics, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:
图1示出了根据本申请实施例的用于生物认证的方法的一个示例性流程图；Fig. 1 shows an exemplary flowchart of a method for biometric authentication according to an embodiment of the present application;
图2示出了根据本申请实施例的判断至少两种生物特征信号是否来自同一真实物体的一个示例性实现流程；FIG. 2 shows an exemplary implementation process of judging whether at least two biometric signals come from the same real object according to an embodiment of the present application;
图3示出了根据本申请实施例的判断至少两种生物特征信号是否来自同一真实物体的另一个示例性实现流程；FIG. 3 shows another exemplary implementation process for judging whether at least two biometric signals come from the same real object according to an embodiment of the present application;
图4示出了根据本申请实施例的用于生物认证的方法的另一个示例性流程图；FIG. 4 shows another exemplary flowchart of a method for biometric authentication according to an embodiment of the present application;
图5示出了根据本申请实施例的基于至少两种生物特征信号进行身份认证或识别的一个示例性实现流程；FIG. 5 shows an exemplary implementation process of identity authentication or identification based on at least two biometric signals according to an embodiment of the present application;
图6示出了根据本申请实施例提供的用于生物认证的装置的一个实施例的结构示意图；FIG. 6 shows a schematic structural diagram of an embodiment of a device for biometric authentication provided according to an embodiment of the present application;
图7示出了根据本申请实施例提供的生物认证系统的一个实施例的结构示意图；以及Figure 7 shows a schematic structural diagram of an embodiment of a biometric authentication system provided according to an embodiment of the present application; and
图8a-图8f示出了根据本申请实施例的生物认证系统的一些示例性实现。Figures 8a-8f illustrate some exemplary implementations of biometric authentication systems according to embodiments of the present application.
具体实施方式detailed description
下面结合附图和实施例对本申请作进一步的详细说明。可以理解的是，此处所描述的具体实施例仅仅用于解释相关发明，而非对该发明的限定。另外还需要说明的是，为了便于描述，附图中仅示出了与有关发明相关的部分。The application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain related inventions, rather than to limit the invention. It should also be noted that, for the convenience of description, only the parts related to the related invention are shown in the drawings.
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and embodiments.
请参考图1，其示出了根据本申请实施例的用于生物认证的方法的一个示例性流程图100。本实施例主要以该方法应用于具有多种生物特征采集和处理功能的身份认证系统中来说明。所述用于生物认证的方法，包括以下步骤：Please refer to FIG. 1 , which shows an exemplary flowchart 100 of a method for biometric authentication according to an embodiment of the present application. This embodiment is mainly described by applying the method to an identity authentication system with multiple biometric features collection and processing functions. The method for biometric authentication includes the following steps:
如图1所示，在步骤110中，接收至少两种生物特征信号。As shown in FIG. 1, in step 110, at least two biometric signals are received.
生物特征是表征生物体的身体或行为特性的特征，例如人脸、指纹、掌纹、血管、虹膜、视网膜、心电、脑电、脉搏、血压、心音、胸或腹运动、人体导电性等。相应地，生物特征信号可以包括人脸图像、指纹图像、掌纹图像、血管图像、虹膜图像、视网膜图像、心电信号、脑电信号、光电容积脉搏波(PPG，PhotoPlethysmoGraphy)信号、血压信号、心音信号、人体调制的电磁波信号、胸或腹运动信号、人体导电性信号以及其他未列举的包含生物特征的信号。本领域技术人员可以理解，上述示例是非穷尽性的，还可以有其他现在已知或者未来开发的各种生物特征信号。Biometrics are features that characterize the physical or behavioral characteristics of an organism, such as face, fingerprint, palm print, blood vessel, iris, retina, ECG, EEG, pulse, blood pressure, heart sound, chest or abdomen movement, human conductivity, etc. . Correspondingly, biometric signals may include face images, fingerprint images, palmprint images, blood vessel images, iris images, retinal images, ECG signals, EEG signals, photoplethysmography (PPG, PhotoPlethysmoGraphy) signals, blood pressure signals, Heart sound signals, electromagnetic wave signals modulated by the human body, chest or abdominal movement signals, human body conductivity signals, and other unlisted signals containing biological characteristics. Those skilled in the art can understand that the above examples are non-exhaustive, and there may also be various biometric signals that are known now or developed in the future.
生物特征信号能够以电信号、声音信号、力信号、电磁信号、图像或视频信号、光信号等多种形式被采集。用于采集生物特征信号的传感器可以具有多种形式。例如，彩色图像传感器，可以用于采集人脸图像、指纹图像、掌纹图像、视网膜图像等；红外图像传感器，可以用于采集血管图像等对红外光源敏感的生物特征的图像；震动传感器，可以用于采集胸或腹运动信号等具有震动特性的信号。其他传感器诸如压力传感器可以用来采集血压信号、胸或腹运动信号等产生压力的生物特征信号。多种生物特征信号可以由多个不同传感器同时分别采集或者由一个集成有多种功能的传感器同时采集。例如，可以分别由彩色图像传感器和光电传感器同时采集人脸图像和PPG信号，或者可以由集成有彩色图像传感器和光电传感器的多功能传感器同时采集人脸图像和PPG信号。Biometric signals can be collected in various forms such as electrical signals, sound signals, force signals, electromagnetic signals, image or video signals, and optical signals. Sensors for collecting biometric signals can take many forms. For example, a color image sensor can be used to collect face images, fingerprint images, palmprint images, retinal images, etc.; an infrared image sensor can be used to collect images of biological characteristics sensitive to infrared light sources such as blood vessel images; a vibration sensor can It is used to collect signals with vibration characteristics such as chest or abdominal motion signals. Other sensors such as pressure sensors can be used to collect pressure-generating biometric signals such as blood pressure signals, chest or abdominal movement signals, and the like. Multiple biometric signals can be collected separately by a plurality of different sensors or simultaneously collected by a sensor integrated with multiple functions. For example, the face image and the PPG signal can be collected simultaneously by the color image sensor and the photoelectric sensor respectively, or the face image and the PPG signal can be collected simultaneously by the multifunctional sensor integrated with the color image sensor and the photoelectric sensor.
在一些实现中，每种生物特征信号的采集都可以持续一预定时间段，以获得随时间变化的连续的生物特征信号或多个在时间上具有关联性的生物特征的离散数据点(例如采集1分钟内的心电信号波形，或者在10秒时间采集多幅人脸图像)。采集到的生物特征信号可以经由信号转换电路进行诸如以下处理之后传送给处理器：将连续的模拟信号(如心电信号、PPG信号等)转换为数字信号，对信号中的噪声进行处理(例如对采集到的虹膜图像去除眼睑、睫毛等图像噪声)，对信号的强度、分布、变化等进行处理(例如对掌纹图像进行归一化处理)。处理器则可以接收经过处理的生物特征信号。In some implementations, the acquisition of each biometric signal can be continued for a predetermined period of time to obtain a continuous biometric signal that varies over time or a plurality of discrete data points of biometric characteristics that are correlated in time (e.g., acquisition ECG signal waveform within 1 minute, or multiple face images collected within 10 seconds). The collected biometric signal can be sent to the processor after being processed by the signal conversion circuit such as: converting continuous analog signals (such as ECG signals, PPG signals, etc.) Remove image noise such as eyelids and eyelashes from the collected iris image), and process the intensity, distribution, and change of the signal (for example, normalize the palmprint image). The processor can then receive the processed biometric signal.
接着，在步骤120中，从至少两种生物特征信号中分别提取同一生理特征。Next, in step 120, the same physiological feature is respectively extracted from at least two biological feature signals.
在本实施例中，生理特征是表征生物体的生理状态的特征，可以是心跳、呼吸、血压、体温等。为了防止使用虚假生物信号进行欺骗，在一些实施例中，提取的生理特征可以是随时间变化的生理特征，例如心跳、呼吸等。如果步骤110中接收到两种以上的生物特征信号，则处理器可以从接收到的每一种生物特征信号中提取同一生理特征。也可以从每一种生物特征信号中提取多种对应的生理特征(例如对于多种生物特征信号同时提取心跳和呼吸)。In this embodiment, the physiological feature is a feature that characterizes the physiological state of the organism, and may be heartbeat, respiration, blood pressure, body temperature, and the like. In order to prevent spoofing with false biological signals, in some embodiments, the extracted physiological features may be physiological features that change over time, such as heartbeat, respiration, and the like. If more than two kinds of biological characteristic signals are received in step 110, the processor may extract the same physiological characteristic from each of the received biological characteristic signals. Multiple corresponding physiological features may also be extracted from each biological feature signal (for example, heartbeat and respiration are simultaneously extracted for multiple biological feature signals).
可以通过多种方法对生理特征进行提取。在一些实现中，可以基于采集到的生物特征信号与生理特征之间的直接映射关系得出生理特征。例如，人体的呼吸率与脉搏率具有比较稳定的比例关系(1:4)，可以根据PPG信号估算出呼吸率。在另一些实现中，可以基于一段时间内的生物特征信号随时间变化的特性与生理特征之间的关系得出生理特征。例如心跳与人脸皮肤下的毛细血管的颜色变化有直接对应的关系，因而可以从一段时间内按时间顺序采集的多幅人脸图像中分析皮肤颜色随时间的变化，并根据该变化得出心跳。Physiological features can be extracted by various methods. In some implementations, the physiological characteristics can be obtained based on the direct mapping relationship between the collected biological characteristic signals and the physiological characteristics. For example, the respiration rate and pulse rate of the human body have a relatively stable proportional relationship (1:4), and the respiration rate can be estimated based on the PPG signal. In some other implementations, the physiological characteristics may be derived based on the relationship between the time-varying characteristics of the biological characteristic signal and the physiological characteristics within a period of time. For example, there is a direct correspondence between the heartbeat and the color change of the capillaries under the skin of the human face. Therefore, the change of skin color over time can be analyzed from multiple face images collected in chronological order over a period of time, and based on this change, heartbeat.
在本实施例中，提取出的生理特征可以以多种形式表示。例如向量、向量组、信号波形等。在一些可选的实现方式中，以向量来表示提取出的生理特征，向量中的每一个元素可以对应于一个信号采集时间，元素的值可以表示生理特征的强度或在其变化过程中所处的位置。具体地，在实际应用中，可以采用例如以下形式的向量来表示生理特征：用一个一维向量来表示心跳，其中N为采样数量，N≥1，a1对应于第一个采样时间点，a1的大小可以表示第一个采样时间点的心电信号强度，依次类推，aN对应于第N个采样时间点，aN的大小可以表示第N个采样时间点的心电信号强度。进一步地，向量中的每个元素也可以对应于一个生理特征发生变化的时刻。例如，表示心跳的向量中的元素a1表示第一个波峰的时间，依次类推，aN表示第N个波峰的时间。在另一些可选的实现方式中，提取出的生理特征可以用一组向量表示，其中每一个向量对应于一个生理特征发生变化的时刻以及变化的程度。例如，可以采用向量组(M≥1)来表示心跳，其第一个向量可以对应于心跳的第一个波峰的发生时刻，a12可以对应于第一个波峰的心电信号强度，依次类推，aM1可以对应于心跳的第M个波峰的发生时刻，aM2可以对应于第M个波峰的心电信号强度。此外，提取出的生理特征还可以以二维坐标系中连续的信号波形来表示，其中，波形的变化趋势可以对应于生理特征的强度的变化趋势。In this embodiment, the extracted physiological features can be expressed in various forms. For example, vectors, groups of vectors, signal waveforms, etc. In some optional implementations, the extracted physiological features are represented by a vector, each element in the vector can correspond to a signal acquisition time, and the value of the element can represent the intensity of the physiological feature or the position in the process of its change. s position. Specifically, in practical applications, a vector of the following form can be used to represent physiological characteristics: a one-dimensional vector to represent the heartbeat, where N is the number of samples, N≥1, a1 corresponds to the first sampling time point, the size of a1 can represent the ECG signal strength at the first sampling time point, and so on, aN corresponds to At the Nth sampling time point, the size of aN can represent the ECG signal strength at the Nth sampling time point. Further, each element in the vector may also correspond to a moment when a physiological feature changes. For example, a vector representing a heartbeat The element a1 in represents the time of the first peak, and so on, and aN represents the time of the Nth peak. In other optional implementation manners, the extracted physiological features may be represented by a set of vectors, where each vector corresponds to the moment when a physiological feature changes and the degree of change. For example, one can take the vector group (M≥1) to represent the heartbeat, the first vector can correspond to the occurrence moment of the first peak of the heartbeat, a12 can correspond to the ECG signal strength of the first peak, and so on, aM1 can correspond to the occurrence moment of the Mth peak of the heartbeat, and aM2 can correspond to ECG signal intensity at the Mth peak. In addition, the extracted physiological features may also be represented by continuous signal waveforms in a two-dimensional coordinate system, wherein the changing trend of the waveform may correspond to the changing trend of the intensity of the physiological features.
然后，在步骤130中，处理生理特征以判断至少两种生物特征信号是否来自同一真实生物体。Then, in step 130, the physiological characteristics are processed to determine whether at least two biological characteristic signals come from the same real organism.
判断至少两种生物特征信号是否来自同一真实物体可以基于来自同一生物体同时采集的至少两种生物特征信号的生理特征表示与来自不同生物体或不同时间采集的至少两种生物特征信号的生理特征表示之间的差异性。换言之，由于来自同一生物体同时采集的至少两种生物特征信号的同一生理特征之间的一致性与来自不同生物体或不同时间采集的至少两种生物特征信号的同一生理特征之间的一致性具有能够区分的差异，因此可以基于上述原理判断接收的至少两种生物特征信号是否来自同一真实物体。例如，不同时刻(例如放松状态的时刻和紧张状态的时刻)基于同一人的人脸图像所获得的心跳频率之间具有较大差异性，基于不同人的人脸图像所获得的心跳频率之间也具有较大差异性。因此，可以基于这种差异性来判断至少两种生物特征信号是否来自同一真实物体。具体地，当这种差异性表现强烈时，可以确认生物特征信号不是来自同一真实生物体，也即采集到的生物特征信号可能是虚假的、仿制或复制的信号。Judging whether the at least two biometric signals come from the same real object can be based on the physiological characteristics of at least two biometric signals collected simultaneously from the same biological body and the physiological characteristics of at least two biometric signals collected from different organisms or at different times Indicates the difference between. In other words, due to the consistency between the same physiological characteristics of at least two biometric signals collected simultaneously from the same organism and the consistency between the same physiological characteristics of at least two biometric signals collected from different organisms or at different times There are differences that can be distinguished, so it can be judged based on the above principles whether the at least two biometric signals received come from the same real object. For example, there is a large difference between the heartbeat frequencies obtained based on the face images of the same person at different moments (such as the moments of the relaxed state and the moments of the tense state), and the heartbeat frequencies obtained based on the face images of different people are different. There are also large differences. Therefore, it can be judged whether at least two biometric signals come from the same real object based on this difference. Specifically, when the difference is strong, it can be confirmed that the biometric signal does not come from the same real organism, that is, the collected biometric signal may be a false, imitated or duplicated signal.
在本实施例中，当生物识别系统被用于认证或识别时，系统可以基于生理特征对被认证者或被识别者进行活体检测，即检测被认证者或被识别者是否为真实生物体，如果传感器采集到的至少两种生物特征信号中包含的同一生理特征表现出一致性，则可以认证被认证者或被识别者为真实的生物体。In this embodiment, when the biometric system is used for authentication or identification, the system can perform liveness detection on the authenticated person or the recognized person based on physiological characteristics, that is, to detect whether the authenticated person or the recognized person is a real organism, If the same physiological feature included in the at least two biological feature signals collected by the sensor shows consistency, the authenticated person or the recognized person can be authenticated as a real biological body.
提取到至少两种生物特征信号中所包含的生理特征之后，可以对生理特征进行如下处理：形式转换，特征分析，相似性分析等。举例而言，当提取到的至少两种生物特征信号具有不同的表示形式时，可以首先将其转换为具有相同的表示形式的信号(例如将从PPG信号中提取的以向量组形式表示的心跳和从人脸图像中提取的以信号波形形式表示的心跳均转换为以一维向量形式表示的心跳)，基于转化后的生理特征确认至少两种生物特征信号是否来自同一真实生物体。又例如，可以对提取到的生理特征进行特征分析，当提取到的至少两种生物信号以向量组形式表示时，可以将向量组转化为矩阵，并通过计算矩阵的特征值或协方差矩阵等来对所提取到的生理特征进行特征分析。After extracting the physiological features contained in the at least two biological feature signals, the following processing may be performed on the physiological features: form conversion, feature analysis, similarity analysis, and the like. For example, when the extracted at least two biometric signals have different representation forms, they may first be converted into signals with the same representation form (for example, the heartbeat expressed in the form of a vector group extracted from the PPG signal and the heartbeat expressed in the form of the signal waveform extracted from the face image are all converted into the heartbeat expressed in the form of a one-dimensional vector), and based on the converted physiological characteristics, it is confirmed whether at least two biometric signals come from the same real organism. For another example, feature analysis can be performed on the extracted physiological features. When the extracted at least two biological signals are represented in the form of a vector group, the vector group can be converted into a matrix, and the eigenvalue or covariance matrix of the matrix can be calculated. To perform feature analysis on the extracted physiological features.
另外，还可以对提取到至少两种生物特征信号中所包含的生理特征做相似性或差异性分析，具体实现方式将会在后面的实施例中详细阐述。In addition, a similarity or difference analysis may also be performed on the physiological features included in the extracted at least two biological feature signals, and the specific implementation manner will be described in detail in the following embodiments.
在本实施例中，可以基于提取的至少两种生物特征信号中的生理特征判断这些信号是否来自于同一个真实的生物体。具体地，如果提取的生理特征具有相同的强度、频率或变化趋势，则可以确认上述至少两种生物特征信号来自于同一生物体。In this embodiment, it may be determined based on the physiological characteristics of at least two extracted biological characteristic signals whether these signals come from the same real biological body. Specifically, if the extracted physiological features have the same intensity, frequency or change trend, it can be confirmed that the above at least two biological feature signals come from the same organism.
本申请的上述实施例通过由处理器接收至少两种生物特征信号，然后从这些生物特征信号中分别提取同一生理特征，之后处理提取出的生理特征以判断这些生物特征信号是否来自同一真实生物体，可以使在不同时间或不同生物体上复制的多种生物特征无法同时使用，显著增加生物特征盗用的难度，实现待认证或待识别的生物体的真实性的判断，提升身份认证系统的安全性。In the above-mentioned embodiments of the present application, the processor receives at least two kinds of biological characteristic signals, and then extracts the same physiological characteristics from these biological characteristic signals, and then processes the extracted physiological characteristics to determine whether these biological characteristic signals come from the same real biological body , which can prevent multiple biometrics copied at different times or on different organisms from being used at the same time, significantly increasing the difficulty of biometric theft, realizing the judgment of the authenticity of the organisms to be authenticated or identified, and improving the security of the identity authentication system sex.
进一步参考图2，其示出了根据本申请实施例的判断至少两种生物特征信号是否来自同一真实物体的一种示例性实现流程200，也即示出了图1的方法步骤130的一个示例性实现的流程图。在本实施例中，通过对提取到至少两种生物特征信号中所包含的生理特征做相似性或差异性分析，做出信号是否来自真实生物体以及是否来自同一生物体的判断。Further referring to FIG. 2 , it shows an exemplary implementation process 200 of judging whether at least two biometric signals come from the same real object according to an embodiment of the present application, that is, an example of method step 130 in FIG. 1 is shown. A flow chart of sexual realization. In this embodiment, by analyzing the similarity or difference of the physiological features included in the extracted at least two biological feature signals, it is judged whether the signals come from a real biological body or from the same biological body.
如图2所示，在步骤210中，基于生理特征的属性来计算至少两种生物特征信号的生理特征之间的一致性度量。As shown in FIG. 2, in step 210, a consistency measure between the physiological characteristics of at least two biometric signals is calculated based on the attributes of the physiological characteristics.
一致性度量是用于表征至少两种生理特征之间的一致性的指标，可以采用多种方法来计算。一致性度量可以用相似度或差异度来表征。举例而言，当提取出的生理特征以信号波形的形式表示时，可以根据生理特征的波形中预定生理现象的采集时间的对应关系来计算一致性度量。在此示例中，预定生理现象可以对应于生理特征的波形中的波峰或波谷。从来自于两种生物信号的生理特征的信号波形中分别检测出波峰或波谷时刻后，根据信号采集的时间，可以得出两组波峰或波谷时刻的对应关系。一致性度量可以用差异度或相似度来表征。当采用差异度表征时，差异度可以表示为生理特征的波形中对应波峰或波谷的采集时间偏差的方差。当采用相似度来表征时，相似度可以表示为上述差异度的倒数。如果两种生物特征信号为同时采集且采集自同一生物体，则对应波峰或波谷时刻的偏差基本恒定，不随时间变化，方差较小，差异度较小，相似度较大，也即一致性较高；而如果两种生物特征信号为不同时刻采集或采集自不同的生物体，则对应波峰或波谷时刻的偏差不稳定，随时间变化剧烈，方差较大，差异度较大，相似度较小，也即一致性较低。The consistency measure is an index used to characterize the consistency between at least two physiological characteristics, and can be calculated by various methods. Consistency measures can be characterized by similarity or difference. For example, when the extracted physiological features are expressed in the form of signal waveforms, the consistency measure may be calculated according to the corresponding relationship between the acquisition times of predetermined physiological phenomena in the waveform of the physiological features. In this example, the predetermined physiological phenomenon may correspond to peaks or troughs in the waveform of the physiological characteristic. After the peak or trough moments are respectively detected from the signal waveforms of the physiological characteristics of the two biological signals, according to the time of signal collection, the corresponding relationship between the two groups of peak or trough moments can be obtained. The consistency measure can be characterized by difference degree or similarity degree. When the degree of difference is used, the degree of difference can be expressed as the variance of the acquisition time deviation corresponding to the peak or trough in the waveform of the physiological feature. When the similarity is used to represent, the similarity can be expressed as the reciprocal of the above-mentioned difference. If the two biometric signals are collected at the same time and from the same organism, the deviation corresponding to the peak or trough moment is basically constant, does not change with time, the variance is small, the difference is small, and the similarity is large, that is, the consistency is high. High; and if the two biometric signals are collected at different times or from different organisms, the deviation corresponding to the peak or trough time is unstable, changes drastically with time, has a large variance, a large degree of difference, and a small degree of similarity , that is, the consistency is low.
可选地，一致性度量也可以采用机器学习的方法进行计算。例如，可以利用回归器来计算一致性度量，其中回归器是经由输入的生理特征数据和设置的一致性度量而训练得到。具体地，可以采用人工产生或采集的生理特征数据和人工设置的一致性度量作为训练样本集产生一个回归器，在训练时对来自同时采集的生物特征信号的生理特征设置较高的一致性度量，而对于不同时采集的或者采集自不同生物体的生物信号的生理特征设置较低的一致性度量。之后可以分别从来自不同生物特征信号的生理特征的向量或信号波形中采样固定长度的数据点，送入该回归器，输出的结果即为一致性度量。Optionally, the consistency measure can also be calculated by using a machine learning method. For example, a regressor may be used to calculate the consistency measure, wherein the regressor is trained through the input physiological characteristic data and the set consistency measure. Specifically, the artificially generated or collected physiological feature data and the manually set consistency measure can be used as a training sample set to generate a regressor, and a higher consistency measure is set for the physiological features from the simultaneously collected biometric signal during training. , and a lower consistency measure is set for the physiological characteristics of biological signals collected at different times or collected from different organisms. Afterwards, fixed-length data points can be sampled from vectors or signal waveforms of physiological characteristics from different biometric signals, and sent to the regressor, and the output result is the consistency measure.
在本实施例的一些可选的实现方式中，该一致性度量可以基于生理特征的以下至少一个属性来计算：时域属性、频域属性和统计属性。In some optional implementation manners of this embodiment, the consistency measure may be calculated based on at least one of the following attributes of physiological features: time domain attributes, frequency domain attributes, and statistical attributes.
在一些可选的实现方式中，时域属性可具体包括生理特征中预定生理现象的发生时刻(如心电图中的波峰时刻、波谷时刻等)、变化时刻、延续时间或生理特征的信号波形；频域属性可以包括生理特征的信号频率或信号频谱。In some optional implementations, the time-domain attribute may specifically include the occurrence moment (such as the peak moment, the trough moment, etc.) of the predetermined physiological phenomenon in the physiological feature, the change moment, the duration, or the signal waveform of the physiological feature; Domain attributes may include signal frequencies or signal spectra of physiological characteristics.
然后，在步骤220中，响应于一致性度量满足预设条件，确认至少两种生物特征信号来自同一真实生物体。Then, in step 220, it is confirmed that at least two biometric signals are from the same real biological body in response to the consistency measure meeting the preset condition.
在本实施例中，根据一致性度量来判断信号是否来自同一真实生物体。一般地，同一时刻同一真实生物体的生理特征之间具有较高的一致性，而不同生物体或者不同时刻采集的生理特征之间具有较低的一致性。如前面所提到的，一致性度量可以用相似度或差异度来表征。相应地，可以为同一时刻同一真实生物体的生理特征的一致性度量设定阈值。如果步骤210中计算得出的一致性度量满足预设条件，例如相似度超过第一预设阈值或者差异度低于第二预设阈值，则可以确认包含该生理特征的生物特征信号来自同一真实生物体，否则，可以确认包含该生理特征的生物特征信号不是来自同一真实生物体或不是同一时刻采集的，其中一个或多个生物特征信号可能是伪造或复制的虚假信号，身份认证系统可以阻止采用这些虚假信号的生物体通过认证。In this embodiment, whether the signals come from the same real organism is judged according to the consistency measure. Generally, there is a high consistency between the physiological characteristics of the same real organism at the same time, and a low consistency between the physiological characteristics collected from different organisms or at different times. As mentioned earlier, the consistency measure can be characterized by similarity or difference. Correspondingly, a threshold can be set for the consistency measure of the physiological characteristics of the same real organism at the same moment. If the consistency measure calculated in step 210 satisfies the preset condition, for example, the similarity exceeds the first preset threshold or the difference is lower than the second preset threshold, then it can be confirmed that the biometric signal containing the physiological feature comes from the same real biometrics, otherwise, it can be confirmed that the biometric signals containing the physiological characteristics are not from the same real biological body or collected at the same time, one or more biometric signals may be fake or copied false signals, the identity authentication system can prevent Organisms that employ these false signals are authenticated.
在本实施的一些可选的实现方式中，阈值的设定可以基于一定数量样本集的训练结果，也可以根据经验值人工设定。其中基于样本集的训练结果设定阈值具体可以如下进行：首先选定同一真实生物体的生理特征的一致性度量样本集和不同生物体或者不同时刻采集的生理特征的一致性度量样本集，在一定范围内选定多个一致性度量值，针对每一个一致性度量值，分别计算该一致性度量值上的同一真实生物体的生理特征的一致性度量的分布密度或分布数量以及不同生物体和不同时刻采集的生理特征的一致性度量的分布密度或分布数量，得出真实生物体的生理特征的一致性度量的分布曲线和不同生物体或不同时刻采集的生理特征的一致性度量的分布曲线，如果两曲线相交，则选择交点所对应的一致性度量作为预设阈值；如果两曲线不相交，则可以在同一真实生物体的生理特征的最小一致性度量值和不同生物体或不同时刻采集的生理特征的最大一致性度量值之间选择一个值作为预设阈值。In some optional implementation manners of this implementation, the setting of the threshold may be based on training results of a certain number of sample sets, or may be manually set according to empirical values. The threshold value setting based on the training results of the sample set can be specifically carried out as follows: firstly, the consistency measurement sample set of the physiological characteristics of the same real organism and the consistency measurement sample set of physiological characteristics collected by different organisms or at different times are selected. Select multiple consistency measures within a certain range, and for each consistency measure, calculate the distribution density or distribution quantity of the consistency measure of the physiological characteristics of the same real organism on the consistency measure and the distribution number of different organisms and the distribution density or distribution quantity of the consistency measure of the physiological characteristics collected at different times to obtain the distribution curve of the consistency measure of the physiological characteristics of the real organism and the distribution of the consistency measure of the physiological characteristics collected at different organisms or at different times curve, if the two curves intersect, the consistency measure corresponding to the intersection point is selected as the preset threshold; if the two curves do not intersect, the minimum consistency measure of the physiological characteristics of the same real organism and different organisms or different moments A value is selected among the maximum consistency measures of the collected physiological characteristics as the preset threshold.
进一步参考图3，其示出了根据本申请实施例的判断至少两种生物特征信号是否来自同一真实物体的另一种示例性实现流程300，也即示出了图1的方法步骤130的另一个示例性实现的流程图。Further referring to FIG. 3 , it shows another exemplary implementation process 300 of judging whether at least two biometric signals come from the same real object according to an embodiment of the present application, that is, another example of step 130 of the method in FIG. 1 is shown. Flowchart of an exemplary implementation.
如图3所示，在步骤310中，基于生理特征的属性，利用分类器对至少两种生物特征信号进行分类。As shown in FIG. 3 , in step 310 , based on the attributes of the physiological features, a classifier is used to classify at least two biological feature signals.
在本实施例中，可以采用分类器对两种生物特征信号进行分类，其中分类器使用两类样本训练得到，第一类样本来自同时采集的同一生物体的生物特征信号，第二类样本来自不同时采集的生物特征信号或者来自不同生物体的生物特征信号。样本大小可以基于训练时间和训练结果精度来确定，输入分类器的特征可以是经过处理的生理特征的属性，例如归一化的时域属性、频域属性等，分类器可以采用支持向量机的算法，在一些实现中，为了提高分类的精度，可以采用级联的弱分类器形成强分类器来对生物特征信号进行分类。In this embodiment, a classifier can be used to classify two types of biometric signals, wherein the classifier is trained using two types of samples, the first type of samples comes from the biometric signals of the same organism collected at the same time, and the second type of samples comes from Biometric signals collected at different times or biometric signals from different organisms. The sample size can be determined based on the training time and the accuracy of the training results. The features of the input classifier can be the attributes of the processed physiological features, such as normalized time domain attributes, frequency domain attributes, etc. The classifier can use the support vector machine Algorithms, in some implementations, in order to improve classification accuracy, cascaded weak classifiers can be used to form a strong classifier to classify biometric signals.
类似地，生理特征的属性可以包含以下至少一项：时域属性、频域属性和统计属性。在一些可选的实现方式中，时域属性可具体包括生理特征中预定生理现象的发生时刻(如心电图中的波峰时刻、波谷时刻等)、变化时刻、延续时间或生理特征的信号波形；频域属性可以包括生理特征的信号频率或信号频谱。Similarly, the attributes of the physiological characteristics may include at least one of the following: time-domain attributes, frequency-domain attributes, and statistical attributes. In some optional implementations, the time-domain attribute may specifically include the occurrence moment (such as the peak moment, the trough moment, etc.) of the predetermined physiological phenomenon in the physiological feature, the change moment, the duration, or the signal waveform of the physiological feature; Domain attributes may include signal frequencies or signal spectra of physiological characteristics.
然后，在步骤320中，根据分类结果确认至少两种生物特征信号是否来自同一真实生物体。Then, in step 320, it is confirmed according to the classification result whether the at least two biometric signals come from the same real organism.
如果分类器输出结果为与训练的第一类样本同一类，则可以确认至少两种生物特征信号来自同一真实生物体，否则，认为至少两种生物特征信号来自不同的生物体或在不同时间采集。If the output result of the classifier is the same as the first type of sample trained, it can be confirmed that at least two biometric signals come from the same real organism, otherwise, at least two biometric signals are considered to be from different organisms or collected at different times .
进一步参考图4，其示出了根据本申请实施例的用于生物认证的方法的另一个示例性流程图400。Further referring to FIG. 4 , it shows another exemplary flowchart 400 of a method for biometric authentication according to an embodiment of the present application.
如图4所示，在步骤410中，接收至少两种生物特征信号。As shown in FIG. 4, in step 410, at least two biometric signals are received.
在本实施例中，处理器可以从传感器接收生物特征信号，这些信号可以由多个传感器通过光学成像、信号检测等方法同时获取，也可以由集成多种功能的传感器经由多种信号获取方法同时获取。In this embodiment, the processor can receive biometric signals from the sensors, and these signals can be simultaneously acquired by multiple sensors through methods such as optical imaging and signal detection, or can be simultaneously acquired by a sensor integrating multiple functions through multiple signal acquisition methods Obtain.
接着，在步骤420中，从至少两种生物特征信号中分别提取同一生理特征。Next, in step 420, the same physiological feature is respectively extracted from at least two biological feature signals.
在本实施例中，对每一种生物特征信号，分别提取同一生理特征，由于生理特征与生物特征信号有关联性(例如心跳频率与皮肤颜色的变化有直接对应的关系)，可以基于这种关联性，分析得出生物特征信号与生理特征的关系，继而根据这种关系从生物特征信号中提取生理特征。提取出的生理特征可以多种形式表示，例如向量、向量组、信号波形等。In this embodiment, for each biological characteristic signal, the same physiological characteristic is extracted respectively. Since the physiological characteristic is related to the biological characteristic signal (for example, there is a direct correspondence between the heartbeat frequency and the change of skin color), it can be based on this Correlation, analyzing the relationship between biometric signals and physiological characteristics, and then extracting physiological characteristics from biometric signals according to this relationship. The extracted physiological features can be expressed in various forms, such as vectors, vector groups, signal waveforms, and so on.
接着，在步骤430中，处理生理特征以判断至少两种生物特征信号是否来自同一真实生物体。Next, in step 430, the physiological characteristics are processed to determine whether the at least two biological characteristic signals come from the same real organism.
提取出生理特征之后，需要对该生理特征进行处理，并判断是否来自同一真实生物体。可以对生理特征进行形式转换和归一化处理，以便基于同样的形式和标准对多种生物特征的真实性进行判断。也可以对生理特征进行特征分析和相似性分析，以确认生物特征信号是否来自同一真实生物体。After the physiological features are extracted, the physiological features need to be processed to determine whether they come from the same real organism. Form conversion and normalization processing can be performed on physiological characteristics, so as to judge the authenticity of multiple biological characteristics based on the same form and standard. Feature analysis and similarity analysis can also be performed on physiological features to confirm whether the biometric signals come from the same real organism.
可以理解，本实施例中步骤410、420和430的实现可以分别与前述实施例中的步骤110、120和130相同，在此不再赘述。It can be understood that the implementation of steps 410, 420, and 430 in this embodiment may be the same as steps 110, 120, and 130 in the foregoing embodiments, and details are not repeated here.
然后，在步骤440中，基于至少两种生物特征信号进行身份认证或识别。具体地，基于此至少两种生物特征信号是否来自同一真实生物体的判断结果来进行身份认证或识别。Then, in step 440, identity authentication or identification is performed based on at least two biometric signals. Specifically, identity authentication or identification is performed based on the judgment result of whether the at least two biometric signals come from the same real biological body.
在本实施例中，将步骤430的判断结果作为身份认证或识别的依据之一，对待认证者或待识别者进行身份认证或识别。生物特征信号中可以包含该生物体区别于其他生物体的特征信息，如身份特征信息。具体地，当步骤430判断至少两种生物特征信号来自于同一真实生物体时，可以基于生物特征信号中的身份特征信息来对待认证者或待识别者进行身份认证或识别。可选地，可以基于多种生物特征信号分别进行身份认证或识别，综合多个认证或识别结果确定被认证者或被识别者的身份。In this embodiment, the judgment result of step 430 is used as one of the basis for identity authentication or identification, and identity authentication or identification is performed on the person to be authenticated or the person to be identified. The biometric signal may contain characteristic information that distinguishes the organism from other organisms, such as identity characteristic information. Specifically, when step 430 judges that at least two biometric signals come from the same real organism, the person to be authenticated or the person to be identified can be authenticated or identified based on the identity information in the biometric signal. Optionally, identity authentication or identification can be performed based on multiple biometric signals, and multiple authentication or identification results can be integrated to determine the identity of the authenticated or identified person.
进一步参考图5，其示出了根据本申请实施例的基于至少两种生物特征信号进行身份认证或识别的一个示例性实现流程500，也即示出了图4的方法步骤440的一个示例性实现的流程图。Further referring to FIG. 5 , it shows an exemplary implementation process 500 of identity authentication or identification based on at least two biometric signals according to an embodiment of the present application, that is, an exemplary method step 440 in FIG. 4 is shown. Implementation flow chart.
如图5所示，在步骤510中，从至少两种生物特征信号的至少一个中提取身份特征信息。As shown in FIG. 5, in step 510, identity feature information is extracted from at least one of at least two biological feature signals.
身份特征信息具有较强的辨识度，以将生物体与其他生物体区别开来。通常情况下，将具有唯一性和稳定性的特征用作身份特征信息。一些可选的身份特征信息可以包括但不限于：人脸图像、指纹图像、掌纹图像、血管图像、虹膜图像、视网膜图像、语音信号、步态特征、签字或笔迹特征、心电信号和脑电信号。本领域技术人员可以理解，上述示例是非穷尽性的，还可以有其他现在已知或者未来开发的各种身份特征信息。在本实施例中，为认证或识别生物体的身份，首先将该生物体的身份特征信息从生物特征信号中提取出来。Identity feature information has a strong degree of recognition to distinguish organisms from other organisms. Usually, unique and stable features are used as identity feature information. Some optional identity feature information may include, but is not limited to: face images, fingerprint images, palm print images, blood vessel images, iris images, retinal images, voice signals, gait features, signature or handwriting features, ECG signals and brain images. electric signal. Those skilled in the art can understand that the above examples are non-exhaustive, and there may also be other various identity feature information that is currently known or will be developed in the future. In this embodiment, in order to authenticate or identify the identity of the biological body, the identity feature information of the biological body is first extracted from the biological feature signal.
有多种方法可以提取身份特征信息。可以直接将接收生物特征信号作为身份特征信息，例如，将人脸图像作为用于身份认证的身份验证信息；也可以对生物特征信号进行特征提取，以提取出的特征表示身份验证信息，例如从连续的行走的人的图像中分析人的行走姿势和步伐频率，并采用滤波等方式提取出步态特征作为身份验证信息；还可以通过分析生物特征信号中的某一特征在一段时间内的变化规律，将变化规律量化表示后作为身份验证信息，例如记录一段时间内的心电信号，将心电信号波峰的发生时刻的变化规律(如发生间隔、每一次发生时与上一次波峰值的差值)量化表示后作为身份验证信息。There are various ways to extract identity characteristic information. The received biometric signal can be directly used as identity feature information, for example, the face image can be used as identity verification information for identity authentication; feature extraction can also be performed on the biometric signal, and the extracted feature can represent identity verification information, such as from Analyze the walking posture and step frequency of people in continuous images of walking people, and use filtering and other methods to extract gait features as identity verification information; you can also analyze the changes of a certain feature in the biometric signal over a period of time Regularity, which quantifies and expresses the change rule as identity verification information, such as recording the ECG signal within a period of time, and records the change rule of the peak occurrence time of the ECG signal (such as the occurrence interval, the difference between each occurrence and the previous peak value) Value) is quantified and expressed as authentication information.
在本实施例的一些可选的实现方式中，可以从其中一种生物特征信号中提取身份特征信息，用于身份认证或识别。在另一些可选的实现方式中，也可以从每一种生物特征信号中都提取身份特征信息，从多种生物特征信号中提取出的可以是同一类型的身份特征信息，也可以是不同类型的身份特征信息。In some optional implementation manners of this embodiment, identity feature information may be extracted from one of the biological feature signals for identity authentication or identification. In other optional implementations, identity feature information can also be extracted from each biometric signal, and the identity feature information extracted from multiple biometric signals can be the same type of identity feature information, or different types identity information.
可以理解，身份特征信息可以与在进行生物体真实性判断中提取出的生物特征相同，也可以不同。当二者相同时，可以省略步骤510，直接利用之前提取的结果。It can be understood that the identity feature information may be the same as or different from the biometric feature extracted in the judgment of the authenticity of the living body. When the two are the same, step 510 can be omitted, and the previously extracted results can be used directly.
接着，在步骤520中，将身份特征信息与已注册的身份特征信息进行匹配。Next, in step 520, the identity feature information is matched with the registered identity feature information.
在本实施例中，可以将提取的身份特征信息与数据库内已注册的身份特征信息进行两种模式的匹配：认证和识别。在认证模式下，需将提取的身份特征信息与数据库中的某一预想生物体的身份特征信息进行匹配；在识别模式下，可以遍历数据库中的所有身份特征信息，查找匹配程度最高的身份特征信息所对应的生物体。In this embodiment, the extracted identity feature information can be matched with the registered identity feature information in the database in two modes: authentication and identification. In the authentication mode, the extracted identity feature information needs to be matched with the identity feature information of an expected organism in the database; in the identification mode, all the identity feature information in the database can be traversed to find the identity feature with the highest matching degree The organism to which the information corresponds.
在对身份特征信息进行匹配之前，还可以对提取到的身份特征信息进行预处理。以虹膜识别为例，可以先从采集到的眼部图像中将虹膜区域分割并归一化，之后还可以对归一化的图像进行去噪和增强等处理，然后采用滤波等方法将虹膜的纹理特征提取出来，之后在虹膜数据库中进行模板匹配。可选地，虹膜特征可以用向量表示，计算提取出的虹膜特征向量与数据库中存储的模板向量之间的相似度量即可实现模板匹配，该相似度量可以包括欧氏距离、汉明距离、平方差、相关系数等。Before matching the identity feature information, the extracted identity feature information can also be preprocessed. Taking iris recognition as an example, the iris area can be segmented and normalized from the collected eye images, and then the normalized image can be denoised and enhanced, and then the iris area can be divided by filtering and other methods. The texture features are extracted, and then template matching is performed in the iris database. Optionally, the iris feature can be represented by a vector, and the template matching can be realized by calculating the similarity measure between the extracted iris feature vector and the template vector stored in the database. The similarity measure can include Euclidean distance, Hamming distance, square difference, correlation coefficient, etc.
然后，在步骤530中，响应于匹配成功并且判断结果确认至少两种生物特征信号来自同一真实生物体，认证或识别生物体的身份。Then, in step 530 , in response to the matching being successful and the judgment result confirming that at least two biometric signals are from the same real biometric, the identity of the biometric is authenticated or recognized.
在本实施例中，如果确认至少两种生物特征信号来自同一真实物体，则允许身份认证系统对生物体的身份进行认证或识别，进而根据步骤520的匹配结果认证或识别生物体的身份。具体地，在认证模式下，如果提取的身份特征信息与数据库中的某一预想生物体的身份特征信息相吻合，则可以确认该生物体即为预想生物体，否则可以确认该生物体为不同于预想生物体的其他生物体；在识别模式下，可以将数据库中与待识别的生物体的身份特征信息匹配程度最高的身份特征信息所对应的身份信息作为待识别生物体的身份信息。In this embodiment, if it is confirmed that at least two biometric signals come from the same real object, the identity authentication system is allowed to authenticate or identify the identity of the organism, and then authenticate or identify the identity of the organism according to the matching result in step 520 . Specifically, in the authentication mode, if the extracted identity feature information matches the identity feature information of an expected organism in the database, it can be confirmed that the organism is the expected organism, otherwise it can be confirmed that the organism is different. In the recognition mode, the identity information corresponding to the identity feature information in the database that most closely matches the identity feature information of the organism to be identified can be used as the identity information of the organism to be identified.
在一些可选的实现方式中，判断结果确认至少两种生物特征信号来自同一真实生物体且步骤510中提取出了多个相同或不同类型的身份特征信息，可以对每一个身份验证信息执行上述步骤502，基于多个匹配结果认证或识别生物体的身份。In some optional implementations, the judgment result confirms that at least two biometric signals come from the same real biological body and multiple identical or different types of identity feature information are extracted in step 510, and the above-mentioned identity verification information can be executed for each identity verification information. Step 502, authenticating or identifying the identity of the living body based on the plurality of matching results.
从图4可以看出，与图1所对应的实施例不同的是，本实施例的示例性流程400多出了基于至少两种生物特征信号进行身份认证或识别的步骤440，通过增加的步骤440，可以在采集到的生物特征信号来自于真实个体时，进一步对该真实个体的身份进行认证和识别。It can be seen from FIG. 4 that, unlike the embodiment corresponding to FIG. 1 , the exemplary process 400 of this embodiment has an additional step 440 of identity authentication or identification based on at least two biometric signals. Through the added steps 440. When the collected biometric signal comes from a real individual, further authenticate and identify the identity of the real individual.
本申请的上述实施例可以实现对虚假生物特征信号的判断以及对真实生物体的身份认证或识别，增强了身份认证系统的安全性。在进行这种身份认证或识别时，要求同时采集至少两种生物特征信号。进一步的，采集可以持续一预定时间段，以便获取随时间变化的生理特征。The above-mentioned embodiments of the present application can realize the judgment of the false biological feature signal and the identity authentication or identification of the real biological body, which enhances the security of the identity authentication system. When performing this identity authentication or identification, it is required to simultaneously collect at least two biometric signals. Further, the collection may last for a predetermined period of time, so as to obtain physiological characteristics that change over time.
进一步参考图6，其示出了根据本申请实施例提供的用于生物认证的装置的一个实施例的结构示意图600。Further referring to FIG. 6 , it shows a schematic structural diagram 600 of an embodiment of a device for biometric authentication provided according to an embodiment of the present application.
如图6所示，用于生物认证的装置600包括：接收单元610，配置用于接收至少两种生物特征信号；提取单元620，配置用于上述至少两种生物特征信号中分别提取同一生理特征；以及判断单元630，配置用于处理生理特征以判断至少两种生物特征信号是否来自同一真实生物体。As shown in FIG. 6 , the device 600 for biometric authentication includes: a receiving unit 610 configured to receive at least two biometric signals; an extraction unit 620 configured to extract the same physiological feature from the at least two biometric signals and a judging unit 630 configured to process physiological features to judge whether at least two biometric signals come from the same real organism.
在本实施例中，接收单元610可以从传感器接收生物特征信号，这些信号可以由多个传感器通过光学成像、信号检测等方法获取，也可以由集成多种功能的传感器经由多种信号获取方法同时获取。然后提取单元620对每一种生物特征信号，分别提取同一生理特征，由于生理特征与生物特征信号具有关联性(例如心跳频率可以反映在皮肤颜色的变化中)，可以基于这种关联性，分析得出生物特征信号与生理特征的关系，继而根据这种关系从生物特征信号中提取生理特征。提取出的生理特征可以多种形式表示，例如向量、向量组、信号波形等。之后判断单元630可以对该生理特征进行处理，并判断是否来自同一真实生物体。判断单元630可以对生理特征进行形式转换和归一化处理，以便基于同样的形式和标准对多种生物特征的真实性进行判断。也可以对生理特征进行特征分析和相似性分析，以确认生物特征信号是否来自同一真实生物体。In this embodiment, the receiving unit 610 can receive biometric signals from sensors, and these signals can be obtained by multiple sensors through optical imaging, signal detection, etc., or can be obtained by sensors integrating multiple functions through multiple signal acquisition methods simultaneously Obtain. Then the extracting unit 620 extracts the same physiological feature for each biological feature signal. Since the physiological feature is related to the biological feature signal (for example, the heartbeat frequency can be reflected in the change of skin color), it can be analyzed based on this correlation. The relationship between the biological characteristic signal and the physiological characteristic is obtained, and then the physiological characteristic is extracted from the biological characteristic signal according to this relationship. The extracted physiological features can be expressed in various forms, such as vectors, vector groups, signal waveforms, and so on. Afterwards, the judging unit 630 can process the physiological characteristics and judge whether they come from the same real organism. The judging unit 630 can perform form conversion and normalization processing on the physiological characteristics, so as to judge the authenticity of multiple biological characteristics based on the same form and standard. Feature analysis and similarity analysis can also be performed on physiological features to confirm whether the biometric signals come from the same real organism.
在一些可选的实现方式中，判断单元630可以基于来自同一生物体同时采集的至少两种生物特征信号的同一生理特征表示之间的一致性与来自不同生物体或不同时间采集的至少两种生物特征信号的同一生理特征表示之间的一致性具有能够区分的的差异来执行判断操作。In some optional implementations, the judging unit 630 may be based on the consistency between the same physiological feature representations of at least two biological feature signals collected simultaneously from the same organism and at least two biological feature signals collected at different times from different organisms. Consistency between the same physiological feature representations of the biometric signal has distinguishable differences to perform judgment operations.
一种可能的判断方式中，判断单元630，具体用于基于所述生理特征的属性来计算所述至少两种生物特征信号的生理特征之间的一致性度量；以及响应于所述一致性度量满足预设条件，确认所述至少两种生物特征信号来自同一真实生物体。In a possible judging manner, the judging unit 630 is specifically configured to calculate a consistency measure between the physiological features of the at least two biological feature signals based on the attributes of the physiological feature; and respond to the consistency measure The preset condition is met, and it is confirmed that the at least two biometric signals come from the same real organism.
一种可能的实现方式中，判断单元630，具体用于根据所述生理特征的波形中预定生理现象的采集时间的对应关系来计算一致性度量。In a possible implementation manner, the judging unit 630 is specifically configured to calculate the consistency measure according to the corresponding relationship between the acquisition time of the predetermined physiological phenomenon in the waveform of the physiological feature.
其中，所述预定生理现象对应于所述生理特征的波形中的波峰或波谷，并且所述一致性度量用差异度或相似度来表征，其中所述差异度表示为所述生理特征的波形中对应波峰或波谷的采集时间偏差的方差，所述相似度表示为所述差异度的倒数。Wherein, the predetermined physiological phenomenon corresponds to a peak or a trough in the waveform of the physiological feature, and the consistency measure is characterized by a degree of difference or a degree of similarity, wherein the degree of difference is expressed as The variance of the acquisition time deviation corresponding to the peak or trough, the similarity is expressed as the reciprocal of the difference.
另一种可能的实现方式中，判断单元630，具体用于利用回归器计算所述一致性度量，其中所述回归器经由输入的生理特征数据和设置的一致性度量而训练得到。In another possible implementation manner, the judging unit 630 is specifically configured to use a regressor to calculate the consistency measure, where the regressor is trained through the input physiological feature data and the set consistency measure.
另一种可能的判断方式中，判断单元630，具体用于基于所述生理特征的属性，利用分类器对所述至少两种生物特征信号进行分类，其中所述分类器使用两类样本训练得到，第一类样本来自同时采集的同一生物体的生物特征信号，第二类样本来自不同时采集的生物特征信号或者来自不同生物体的生物特征信号；以及根据分类结果确认所述至少两种生物特征信号是否来自同一真实生物体。In another possible judging manner, the judging unit 630 is specifically configured to use a classifier to classify the at least two kinds of biological characteristic signals based on the attributes of the physiological characteristics, wherein the classifier is trained using two types of samples to obtain , the first type of samples come from biometric signals of the same organism collected at the same time, the second type of samples come from biometric signals collected at different times or from different organisms; and confirming that the at least two types of biological Whether the characteristic signal is from the same real organism.
在本实施例的一些可选的实现方式中，用于生物认证的装置600还可以包括认证与识别单元(未示出)，配置用于基于至少两种生物特征信号进行身份认证或识别，其中将至少两种生物特征信号是否来自同一真实生物体的判断结果作为身份认证或识别的依据之一。In some optional implementations of this embodiment, the device 600 for biometric authentication may further include an authentication and identification unit (not shown), configured to perform identity authentication or identification based on at least two biometric signals, wherein The result of judging whether at least two biometric signals come from the same real biological body is used as one of the basis for identity authentication or identification.
具体的，认证与识别单元，具体用于将从所述至少两种生物特征信号中提取出的身份特征信息与已注册的身份特征信息进行匹配；响应于匹配成功并且所述判断结果确认所述至少两种生物特征信号来自同一真实生物体，认证或识别所述生物体的身份。Specifically, the authentication and identification unit is specifically configured to match the identity feature information extracted from the at least two biological feature signals with the registered identity feature information; At least two biometric signals originate from the same real organism, authenticating or identifying the identity of said organism.
描述于本申请实施例中所涉及到的单元可以通过软件的方式实现，也可以通过硬件的方式来实现。所描述的单元也可以设置在处理器中，例如，可以描述为：一种处理器包括接收单元，提取单元和判断单元。其中，这些单元的名称在某种情况下并不构成对该单元本身的限定，例如，接收单元还可以被描述为“用于接收至少两种生物特征信号的单元”。The units involved in the embodiments described in the present application may be implemented by means of software or by means of hardware. The described units can also be set in a processor, for example, it can be described as: a processor includes a receiving unit, an extracting unit and a judging unit. Wherein, the names of these units do not constitute a limitation on the unit itself under certain circumstances, for example, the receiving unit may also be described as "a unit for receiving at least two biometric signals".
本申请提供的用于生物认证的方法可以用于多种复合生物特征的身份认证系统中，因此，另一方面，本申请还提供了一种生物认证系统。图7示出了根据本申请实施例提供的生物认证系统的一种实施例的结构示意图。如图7所示，生物认证系统700包括传感器710和处理器730。传感器710配置用于采集至少两种生物特征信号；处理器730配置用于从传感器接收至少两种生物特征信号，从至少两种生物特征信号中分别提取同一生理特征，以及处理生理特征以判断所述至少两种生物特征信号是否来自同一真实生物体。在一些实施例中，传感器配置用于同时采集至少两种生物特征信号。The method for biometric authentication provided in this application can be used in multiple identity authentication systems with composite biometric features. Therefore, on the other hand, this application also provides a biometric authentication system. Fig. 7 shows a schematic structural diagram of an embodiment of a biometric authentication system provided according to an embodiment of the present application. As shown in FIG. 7 , the biometric authentication system 700 includes a sensor 710 and a processor 730 . The sensor 710 is configured to collect at least two biological characteristic signals; the processor 730 is configured to receive at least two biological characteristic signals from the sensor, extract the same physiological characteristic from the at least two biological characteristic signals, and process the physiological characteristics to determine the Whether the at least two biometric signals come from the same real organism. In some embodiments, the sensor is configured to acquire at least two biometric signals simultaneously.
上述生物认证系统700还可以包含信号转换电路720。The above biometric authentication system 700 may also include a signal conversion circuit 720 .
在一些实现中，传感器710采集生物特征信号时可以持续一预定时间段，以获得随时间变化的连续的生物特征信号或多个在时间上具有关联性的生物特征的离散数据点。采集到的生物特征信号可以经由信号转换电路720进行诸如以下处理，转换为可以被处理器730处理的形式，然后传送给处理器730：将连续的模拟信号转换为数字信号，对信号中的噪声进行处理，对信号的强度、分布、变化等进行处理。In some implementations, the sensor 710 may collect the biometric signal for a predetermined period of time to obtain a continuous biometric signal that varies over time or a plurality of discrete data points of biometric characteristics that are temporally correlated. The collected biometric signal can be processed by the signal conversion circuit 720 such as the following, converted into a form that can be processed by the processor 730, and then sent to the processor 730: converting the continuous analog signal into a digital signal, eliminating the noise in the signal Perform processing, and process the strength, distribution, and changes of the signal.
上述生物认证系统700还可以包含存储器740，配置用于保存对生物特征信号进行处理过程中使用的指令、参数、数据等，或者用于记录系统运行时得到的数据和结果。The above-mentioned biometric authentication system 700 may also include a memory 740 configured to store instructions, parameters, data, etc. used in processing the biometric signal, or to record data and results obtained during system operation.
上述生物认证系统700还可以包含输出装置750，用来输出处理器处理得到的结果。例如显示使用说明、显示信号是否真实、显示身份认证或识别的结果，操作其他设备、软件等。The above-mentioned biometric authentication system 700 may further include an output device 750 for outputting the result obtained by the processor. For example, displaying instructions for use, displaying whether the signal is authentic, displaying the result of identity authentication or identification, operating other equipment, software, etc.
上述生物认证系统700可以在各种需要进行身份认证或识别的设备上实现。可以用于判断被认证者是否为真实生物体，也可以用于验证被认证者与注册者是否为同一生物体，也可以用于从多个注册者中识别出某个生物体。The above-mentioned biometric authentication system 700 can be implemented on various devices that need to perform identity authentication or identification. It can be used to judge whether the authenticated person is a real organism, or to verify whether the authenticated person and the registrant are the same organism, or to identify a certain organism from multiple registrants.
作为示例，图8a-图8f示出了根据本申请实施例的生物认证系统的一些示例性实现。As an example, Figures 8a-8f show some exemplary implementations of biometric authentication systems according to embodiments of the present application.
如图8a所示，其示出了基于心电信号和PPG信号的具有身份认证功能的智能手表810。手表的正面和背面分别安装有心电图仪(ECG，ElectroCardioGraphy)电极812和814，用于采集心电信号；PPG传感器813用于采集PPG信号。处理器可以接收采集到的心电信号和PPG信号，并从中提取心跳或呼吸特征，然后对心跳或呼吸特征进行分析，根据提取的心跳或呼吸特征的一致性判断是否为同一时刻采集的同一用户的信号，之后还可以对用户身份进行验证，将判断和验证的结果显示在屏幕811上。As shown in Fig. 8a, it shows a smart watch 810 with identity authentication function based on ECG signals and PPG signals. Electrocardiograph (ECG, ElectroCardioGraphy) electrodes 812 and 814 are respectively installed on the front and back of the watch for collecting ECG signals; the PPG sensor 813 is used for collecting PPG signals. The processor can receive the collected ECG signal and PPG signal, and extract the heartbeat or breathing feature from it, and then analyze the heartbeat or breathing feature, and judge whether it is the same user collected at the same time according to the consistency of the extracted heartbeat or breathing feature After that, the identity of the user can be verified, and the results of the judgment and verification can be displayed on the screen 811.
图8b示出了基于人脸图像和心电信号的具有身份认证功能的智能手机820。如图8b所示，屏幕821用于显示采集的信号和处理结果；摄像头822可以用于采集人脸图像；ECG电极823和824可以采集心电信号。手机的处理器可以接收采集到的人脸图像和心电信号，并从中提取心跳特征，然后可以对提取到的心跳特征进行处理，并基于处理后的心跳特征对用户的身份进行真实性判断以及进一步的验证。Fig. 8b shows a smart phone 820 with identity authentication function based on face images and ECG signals. As shown in Fig. 8b, the screen 821 is used to display the collected signals and processing results; the camera 822 can be used to collect face images; the ECG electrodes 823 and 824 can collect electrocardiographic signals. The processor of the mobile phone can receive the collected face images and ECG signals, and extract heartbeat features from them, then process the extracted heartbeat features, and judge the authenticity of the user's identity based on the processed heartbeat features and Further verification.
图8c示出了基于人脸图像、心电信号和拉力传感器的具有身份认证功能的汽车830。如图8c所示，汽车后视镜上的摄像头831用于采集人脸图像；方向盘上安装有ECG电极832，用于采集心电信号，拉力传感器833安装在安全带上，也可以用于采集心电信号。处理器可以接收采集到的信号，并分别从每一种信号中提取心跳或呼吸特征，然后可以对提取到的心跳或呼吸特征进行处理，并基于处理后的心跳或呼吸特征一致性判断这些信号是否来自同一真实生物体。Fig. 8c shows a car 830 with identity authentication function based on face image, ECG signal and tension sensor. As shown in Figure 8c, the camera 831 on the car rearview mirror is used to collect face images; the ECG electrode 832 is installed on the steering wheel to collect ECG signals, and the tension sensor 833 is installed on the seat belt, which can also be used to collect ECG. The processor can receive the collected signals, and extract heartbeat or breathing features from each signal, and then process the extracted heartbeat or breathing features, and judge these signals based on the consistency of the processed heartbeat or breathing features whether from the same real organism.
图8d示出了基于人脸图像和指纹(或掌纹、血管)图像的具有身份认证功能的门锁系统840。如图8d所示，门上方的摄像头841采集人脸图像，门锁上安装有传感器842，可采集指纹、静脉、掌纹图像。Fig. 8d shows a door lock system 840 with identity authentication function based on face image and fingerprint (or palm print, blood vessel) image. As shown in Figure 8d, a camera 841 above the door collects face images, and a sensor 842 is installed on the door lock to collect images of fingerprints, veins, and palm prints.
图8e示出了基于人脸图像和心跳传感器的智能手机850。如图8e所示，手机摄像头852用于采集人脸图像；耳机上安装有脉搏探测器853。当用户进行身份认证时，手机摄像头852的图像传感器采集人脸图像，处理器可以接收采集到的人脸图像，从中提取心跳特征；位于耳塞上的脉搏传感器853获得心跳特征，处理器对两个心跳特征进行分析，并做出待认证生物体的真实性判断。所获取的图像和脉搏信号可以显示在屏幕851上。Figure 8e shows a smartphone 850 based on a face image and a heartbeat sensor. As shown in Figure 8e, the mobile phone camera 852 is used to collect face images; a pulse detector 853 is installed on the earphone. When the user performs identity authentication, the image sensor of the mobile phone camera 852 collects face images, and the processor can receive the collected face images to extract heartbeat features; The heartbeat characteristics are analyzed, and the authenticity judgment of the organism to be authenticated is made. The acquired images and pulse signals can be displayed on the screen 851 .
图8f示出了基于心电信号和脑电信号的具有身份认证功能的手环和耳机860。当用户进行身份认证时，位于手环上的ECG传感器861采集心电信号；耳机上安装有EEG传感器862，用于采集脑电信号。当用户进行身份认证时，处理器可以接收心电信号和脑电信号，然后从心电信号和脑电信号中提取出心跳特征，并通过无线通信的方式将提取出的心跳特征传送到同一个设备上，之后该设备判断提取出的心跳是否来自同一真实生物体并可以进一步对该生物体进行身份验证或识别。Fig. 8f shows a wristband and earphone 860 with identity authentication function based on electrocardiographic signals and electroencephalogram signals. When the user performs identity authentication, the ECG sensor 861 on the bracelet collects electrocardiographic signals; the earphone is equipped with an EEG sensor 862 for collecting electroencephalogram signals. When the user is authenticated, the processor can receive the ECG signal and the EEG signal, and then extract the heartbeat characteristics from the ECG signal and the EEG signal, and transmit the extracted heartbeat characteristics to the same computer through wireless communication. After that, the device judges whether the extracted heartbeat comes from the same real biological body and can further authenticate or identify the biological body.
作为另一方面，本申请还提供了一种计算机可读存储介质，该计算机可读存储介质可以是上述实施例中所述装置中所包含的计算机可读存储介质；也可以是单独存在，未装配入终端中的计算机可读存储介质。所述计算机可读存储介质存储有一个或者一个以上程序，所述程序被一个或者一个以上的处理器用来执行描述于本申请的用于生物认证的方法。As another aspect, the present application also provides a computer-readable storage medium, which may be the computer-readable storage medium contained in the device described in the above-mentioned embodiments; A computer-readable storage medium assembled in a terminal. The computer-readable storage medium stores one or more programs, and the programs are used by one or more processors to execute the method for biometric authentication described in this application.
以上描述仅为本申请的较佳实施例以及对所运用技术原理的说明。本领域技术人员应当理解，本申请中所涉及的发明范围，并不限于上述技术特征的特定组合而成的技术方案，同时也应涵盖在不脱离所述发明构思的情况下，由上述技术特征或其等同特征进行任意组合而形成的其它技术方案。例如上述特征与本申请中公开的(但不限于)具有类似功能的技术特征进行互相替换而形成的技术方案。The above description is only a preferred embodiment of the present application and an illustration of the applied technical principle. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solution formed by the specific combination of the above-mentioned technical features, but should also cover the technical solution formed by the above-mentioned technical features without departing from the inventive concept. Other technical solutions formed by any combination of or equivalent features thereof. For example, a technical solution formed by replacing the above-mentioned features with technical features with similar functions disclosed in (but not limited to) this application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410816701.2ACN105787420B (en) | 2014-12-24 | 2014-12-24 | Method, device and biometric authentication system for biometric authentication |
| KR1020150066256AKR102367481B1 (en) | 2014-12-24 | 2015-05-12 | Method and device to authenticate living body |
| US14/884,004US10154818B2 (en) | 2014-12-24 | 2015-10-15 | Biometric authentication method and apparatus |
| EP15193235.7AEP3037036B1 (en) | 2014-12-24 | 2015-11-05 | Biometric authentication method and apparatus |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410816701.2ACN105787420B (en) | 2014-12-24 | 2014-12-24 | Method, device and biometric authentication system for biometric authentication |
| Publication Number | Publication Date |
|---|---|
| CN105787420Atrue CN105787420A (en) | 2016-07-20 |
| CN105787420B CN105787420B (en) | 2020-07-14 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410816701.2AActiveCN105787420B (en) | 2014-12-24 | 2014-12-24 | Method, device and biometric authentication system for biometric authentication |
| Country | Link |
|---|---|
| KR (1) | KR102367481B1 (en) |
| CN (1) | CN105787420B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106971157A (en)* | 2017-03-22 | 2017-07-21 | 重庆科技学院 | Fingerprint and face coupled identification method based on multiple linear regression associative memory model |
| CN107088069A (en)* | 2017-03-29 | 2017-08-25 | 西安电子科技大学 | Personal identification method based on human body PPG signal subsections |
| CN107277265A (en)* | 2017-07-13 | 2017-10-20 | 广东欧珀移动通信有限公司 | Unlock control method and related products |
| CN107330404A (en)* | 2017-06-30 | 2017-11-07 | 重庆科技学院 | Personal identification method based on cell neural network autoassociative memories model |
| CN107736874A (en)* | 2017-08-25 | 2018-02-27 | 百度在线网络技术（北京）有限公司 | A kind of method, apparatus of In vivo detection, equipment and computer-readable storage medium |
| CN108029665A (en)* | 2017-12-07 | 2018-05-15 | 湖州华科信息咨询有限公司 | A kind of safe insect catching method and device based on robot arm |
| WO2018201683A1 (en)* | 2017-05-03 | 2018-11-08 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for identifying homology of physiological signals |
| CN109145548A (en)* | 2017-06-19 | 2019-01-04 | 韩国网络振兴院 | User authentication device and method and computer readable recording medium based on multi-mode |
| CN109286499A (en)* | 2018-09-21 | 2019-01-29 | 武汉大学 | A Presence Authentication Method Based on Behavior Characteristics |
| CN109409249A (en)* | 2018-09-30 | 2019-03-01 | 联想(北京)有限公司 | Information processing method and electronic equipment |
| CN109784124A (en)* | 2017-11-10 | 2019-05-21 | 北京嘀嘀无限科技发展有限公司 | A kind of determination method of vivo identification, decision-making system and computer installation |
| CN110664383A (en)* | 2019-09-27 | 2020-01-10 | 北京小米移动软件有限公司 | Thermometer, storage method and device for body temperature information |
| CN111758096A (en)* | 2018-01-19 | 2020-10-09 | 奈米公司 | Live user authentication device, system and method |
| CN112364808A (en)* | 2020-11-24 | 2021-02-12 | 哈尔滨工业大学 | Living body identity authentication method based on FMCW radar and face tracking identification |
| CN109844746B (en)* | 2017-03-13 | 2021-05-04 | 华为技术有限公司 | A verification method and verification device |
| US20210251567A1 (en)* | 2020-02-19 | 2021-08-19 | Faceheart Inc. | Device and method for liveness detection |
| CN113273961A (en)* | 2020-02-19 | 2021-08-20 | 钜怡智慧股份有限公司 | Living body detection device and method |
| CN114743218A (en)* | 2022-04-08 | 2022-07-12 | 中国银行股份有限公司 | An identity verification method and system |
| US11398111B2 (en) | 2020-03-02 | 2022-07-26 | Egis Technology Inc. | Integrated spectrum sensing device for real-finger judgement and sensing method |
| CN117037340A (en)* | 2023-09-06 | 2023-11-10 | 东莞市安邦德智能锁具科技有限公司 | Intelligent lock control system based on data identification |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102299847B1 (en)* | 2017-06-26 | 2021-09-08 | 삼성전자주식회사 | Face verifying method and apparatus |
| KR101986246B1 (en)* | 2017-10-23 | 2019-06-05 | 한국인터넷진흥원 | User authentication apparatus using bio-signal and method of the same |
| KR102151424B1 (en)* | 2018-09-19 | 2020-09-03 | 상명대학교산학협력단 | Apparatus and method for discriminating fake finger vein |
| US10817246B2 (en)* | 2018-12-28 | 2020-10-27 | Baidu Usa Llc | Deactivating a display of a smart display device based on a sound-based mechanism |
| KR102663389B1 (en)* | 2019-03-26 | 2024-05-08 | 엘지전자 주식회사 | Electronic device and its control method |
| KR102806305B1 (en) | 2019-07-04 | 2025-05-09 | 삼성전자주식회사 | Apparatus and method for calibrating bio-information estimation model and, apparatus for estimating bio-information |
| KR102798960B1 (en)* | 2023-05-22 | 2025-04-21 | 재단법인대구경북과학기술원 | System for identification of human based on biometric signal and movement and method thereof |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030135097A1 (en)* | 2001-06-25 | 2003-07-17 | Science Applications International Corporation | Identification by analysis of physiometric variation |
| CN1672357A (en)* | 2002-07-29 | 2005-09-21 | C－西格雷特有限公司 | Method and device for electrical biometric identification |
| CN103761465A (en)* | 2014-02-14 | 2014-04-30 | 上海云亨科技有限公司 | Method and device for identity authentication |
| US20140188770A1 (en)* | 2011-05-10 | 2014-07-03 | Foteini Agrafioti | System and method for enabling continuous or instantaneous identity recognition based on physiological biometric signals |
| CN104054038A (en)* | 2011-10-26 | 2014-09-17 | 索尼电脑娱乐公司 | Individual Body Discrimination Device And Individual Body Discrimination Method |
| WO2014157896A1 (en)* | 2013-03-24 | 2014-10-02 | 서울대학교산학협력단 | Film-type device for measuring biomedical signal, and blood pressure measurement device, cardiopulmonary endurance estimation device, and individual certification device using same |
| US20140304792A1 (en)* | 2006-01-30 | 2014-10-09 | Adidas Ag | System and method for identity confirmation using physiologic biometrics to determine a physiologic fingerprint |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8285010B2 (en)* | 2007-03-21 | 2012-10-09 | Lumidigm, Inc. | Biometrics based on locally consistent features |
| US8542879B1 (en)* | 2012-06-26 | 2013-09-24 | Google Inc. | Facial recognition |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030135097A1 (en)* | 2001-06-25 | 2003-07-17 | Science Applications International Corporation | Identification by analysis of physiometric variation |
| CN1672357A (en)* | 2002-07-29 | 2005-09-21 | C－西格雷特有限公司 | Method and device for electrical biometric identification |
| US20140304792A1 (en)* | 2006-01-30 | 2014-10-09 | Adidas Ag | System and method for identity confirmation using physiologic biometrics to determine a physiologic fingerprint |
| US20140188770A1 (en)* | 2011-05-10 | 2014-07-03 | Foteini Agrafioti | System and method for enabling continuous or instantaneous identity recognition based on physiological biometric signals |
| CN104054038A (en)* | 2011-10-26 | 2014-09-17 | 索尼电脑娱乐公司 | Individual Body Discrimination Device And Individual Body Discrimination Method |
| WO2014157896A1 (en)* | 2013-03-24 | 2014-10-02 | 서울대학교산학협력단 | Film-type device for measuring biomedical signal, and blood pressure measurement device, cardiopulmonary endurance estimation device, and individual certification device using same |
| CN103761465A (en)* | 2014-02-14 | 2014-04-30 | 上海云亨科技有限公司 | Method and device for identity authentication |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109844746B (en)* | 2017-03-13 | 2021-05-04 | 华为技术有限公司 | A verification method and verification device |
| CN106971157A (en)* | 2017-03-22 | 2017-07-21 | 重庆科技学院 | Fingerprint and face coupled identification method based on multiple linear regression associative memory model |
| CN107088069A (en)* | 2017-03-29 | 2017-08-25 | 西安电子科技大学 | Personal identification method based on human body PPG signal subsections |
| CN110312466A (en)* | 2017-05-03 | 2019-10-08 | 深圳迈瑞生物医疗电子股份有限公司 | Homology identification method and device for physiological signals |
| US11457873B2 (en) | 2017-05-03 | 2022-10-04 | Shenzhen Mindray Bio-Medical Electronics Co., Ltd. | Method and apparatus for identifying homology of physiological signals |
| WO2018201683A1 (en)* | 2017-05-03 | 2018-11-08 | 深圳迈瑞生物医疗电子股份有限公司 | Method and device for identifying homology of physiological signals |
| CN108784650A (en)* | 2017-05-03 | 2018-11-13 | 深圳迈瑞生物医疗电子股份有限公司 | The homology recognition methods of physiological signal and device |
| CN109145548A (en)* | 2017-06-19 | 2019-01-04 | 韩国网络振兴院 | User authentication device and method and computer readable recording medium based on multi-mode |
| CN107330404A (en)* | 2017-06-30 | 2017-11-07 | 重庆科技学院 | Personal identification method based on cell neural network autoassociative memories model |
| CN107277265A (en)* | 2017-07-13 | 2017-10-20 | 广东欧珀移动通信有限公司 | Unlock control method and related products |
| US11147474B2 (en) | 2017-08-25 | 2021-10-19 | Baidu Online Network Technology (Beijing) Co., Ltd. | Living body detecting method and apparatus, device and computer storage medium |
| CN107736874A (en)* | 2017-08-25 | 2018-02-27 | 百度在线网络技术（北京）有限公司 | A kind of method, apparatus of In vivo detection, equipment and computer-readable storage medium |
| CN109784124A (en)* | 2017-11-10 | 2019-05-21 | 北京嘀嘀无限科技发展有限公司 | A kind of determination method of vivo identification, decision-making system and computer installation |
| CN108029665A (en)* | 2017-12-07 | 2018-05-15 | 湖州华科信息咨询有限公司 | A kind of safe insect catching method and device based on robot arm |
| CN108029665B (en)* | 2017-12-07 | 2020-11-06 | 东阳市聚冉电子科技有限公司 | A method and device for safely trapping insects based on a robotic arm |
| CN111758096A (en)* | 2018-01-19 | 2020-10-09 | 奈米公司 | Live user authentication device, system and method |
| CN109286499B (en)* | 2018-09-21 | 2020-08-07 | 武汉大学 | A Presence Authentication Method Based on Behavior Characteristics |
| CN109286499A (en)* | 2018-09-21 | 2019-01-29 | 武汉大学 | A Presence Authentication Method Based on Behavior Characteristics |
| CN109409249A (en)* | 2018-09-30 | 2019-03-01 | 联想(北京)有限公司 | Information processing method and electronic equipment |
| CN110664383A (en)* | 2019-09-27 | 2020-01-10 | 北京小米移动软件有限公司 | Thermometer, storage method and device for body temperature information |
| US20210251567A1 (en)* | 2020-02-19 | 2021-08-19 | Faceheart Inc. | Device and method for liveness detection |
| CN113273961A (en)* | 2020-02-19 | 2021-08-20 | 钜怡智慧股份有限公司 | Living body detection device and method |
| US11517253B2 (en)* | 2020-02-19 | 2022-12-06 | Faceheart Inc. | Device and method for liveness detection |
| CN113273961B (en)* | 2020-02-19 | 2024-05-24 | 钜怡智慧股份有限公司 | Living body detection device and method |
| US11398111B2 (en) | 2020-03-02 | 2022-07-26 | Egis Technology Inc. | Integrated spectrum sensing device for real-finger judgement and sensing method |
| CN112364808A (en)* | 2020-11-24 | 2021-02-12 | 哈尔滨工业大学 | Living body identity authentication method based on FMCW radar and face tracking identification |
| CN114743218A (en)* | 2022-04-08 | 2022-07-12 | 中国银行股份有限公司 | An identity verification method and system |
| CN117037340A (en)* | 2023-09-06 | 2023-11-10 | 东莞市安邦德智能锁具科技有限公司 | Intelligent lock control system based on data identification |
| CN117037340B (en)* | 2023-09-06 | 2024-04-12 | 东莞市安邦德智能锁具科技有限公司 | Intelligent lock control system based on data identification |
| Publication number | Publication date |
|---|---|
| KR20160078208A (en) | 2016-07-04 |
| CN105787420B (en) | 2020-07-14 |
| KR102367481B1 (en) | 2022-02-24 |
| Publication | Publication Date | Title |
|---|---|---|
| CN105787420B (en) | Method, device and biometric authentication system for biometric authentication | |
| US10154818B2 (en) | Biometric authentication method and apparatus | |
| Abdeldayem et al. | A novel approach for ECG-based human identification using spectral correlation and deep learning | |
| CA2835460C (en) | System and method for enabling continuous or instantaneous identity recognition based on physiological biometric signals | |
| EP4024351B1 (en) | Biometric system | |
| US9946942B2 (en) | Method, apparatus and system for biometric identification | |
| Abo-Zahhad et al. | Biometric authentication based on PCG and ECG signals: present status and future directions | |
| CN104573458B (en) | A kind of personal identification method based on electrocardiosignal, apparatus and system | |
| US9603556B2 (en) | Device and method for continuous biometric recognition based on electrocardiographic signals | |
| Chauhan et al. | A survey of emerging biometric modalities | |
| Islam et al. | Biometric template extraction from a heartbeat signal captured from fingers | |
| WO2018152711A1 (en) | Electrocardiographic authentication-based door control system and authentication method therefor | |
| CN101558997A (en) | EEG signal recognition method based on second-order blind recognition | |
| WO2017187422A1 (en) | Biometric method and device for identifying a person through an electrocardiogram (ecg) waveform | |
| Sasikala et al. | Identification of individuals using electrocardiogram | |
| KR102021491B1 (en) | Apparatus and method for user authentication | |
| KR101587874B1 (en) | A biometric recognition apparatus based on single lead electrocardiography and a method thereof | |
| Pouryayevali | ECG biometrics: new algorithm and multimodal biometric system | |
| Chen et al. | Finger ECG-based authentication for healthcare data security using artificial neural network | |
| Shahid et al. | A survey on AI-based ECG, PPG, and PCG signals based biometric authentication system | |
| Hendrawan et al. | Identification of optimum segment in single channel EEG biometric system | |
| KR101281852B1 (en) | Biometric authentication device and method using brain signal | |
| Yeen et al. | Development of heartbeat based biometric system using wavelet transform | |
| Sahadat et al. | Hardware-efficient robust biometric identification from 0.58 second template and 12 features of limb (Lead I) ECG signal using logistic regression classifier | |
| WO2019153074A1 (en) | System and method for liveness detection and automatic template updating using fusion of medical and non-medical biometrics |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |