CN114550314A - A biometric identification method and device - Google Patents
A biometric identification method and deviceDownload PDFInfo
- Publication number
- CN114550314A CN114550314ACN202210049340.8ACN202210049340ACN114550314ACN 114550314 ACN114550314 ACN 114550314ACN 202210049340 ACN202210049340 ACN 202210049340ACN 114550314 ACN114550314 ACN 114550314A
- Authority
- CN
- China
- Prior art keywords
- recognition
- layered
- identification
- image
- image block
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Collating Specific Patterns (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及生物特征识别技术领域，特别是涉及一种生物特征识别方法及装置。The present invention relates to the technical field of biometric identification, in particular to a biometric identification method and device.
背景技术Background technique
人体的生物特征由于具有唯一性、隐私性、不可更改性等特性，因此广泛应用于身份识别领域。The biometrics of the human body are widely used in the field of identity recognition due to their uniqueness, privacy, and immutability.
目前，在对人体的生物特征进行生物特征识别时，通常采用的生物特征识别方法包括如下两种：第一种，使用生物特征识别模型对生物特征的单张图像进行生物特征识别，此种方式，虽然生物特征识别速度较快，但是由于图像采集可能受到诸如光照、生物特征采集角度等因素的影响，造成图像的质量较差，从而导致生物特征识别的准确度较低。第二种，对单张图像进行分块，然后使用生物特征识别模型对图像分块进行生物特征识别。此种方式虽然生物特征识别相对于第一种方法的准确度有所提高，但是由于图像分块的数量众多，严重影响了生物特征识别的速度，导致生物特征识别的速度较低。At present, when performing biometric identification on the biometrics of the human body, the biometric identification methods usually used include the following two: First, the biometric identification model is used to perform biometric identification on a single image of the biometric. This method , Although the biometric recognition speed is fast, because the image acquisition may be affected by factors such as illumination, biometric acquisition angle, etc., the quality of the image is poor, resulting in a low accuracy of biometric recognition. The second is to block a single image, and then use the biometric recognition model to perform biometric identification of the image blocks. Although the accuracy of biometric identification in this method is improved compared to the first method, due to the large number of image blocks, the speed of biometric identification is seriously affected, resulting in a low speed of biometric identification.
可见，现有的生物特征识别方法存在生物特征识别准确度低和生物特征识别速度低的缺陷。It can be seen that the existing biometric identification methods have the defects of low biometric identification accuracy and low biometric identification speed.
发明内容SUMMARY OF THE INVENTION
有鉴于此，本发明提出了一种生物特征识别方法及装置，主要目的在于在提高生物特征识别准确度的同时，提高生物特征识别的识别速度。In view of this, the present invention proposes a method and device for biometric identification, the main purpose of which is to improve the identification speed of biometric identification while improving the accuracy of biometric identification.
第一方面，本发明提供了一种生物特征识别方法，该方法包括：In a first aspect, the present invention provides a biometric identification method, the method comprising:
对目标生物特征的至少两个图像进行分块处理，形成各所述图像各自的图像分块组，其中，各所述图像具有不同的模态；Perform block processing on at least two images of the target biometric feature to form respective image block groups for each of the images, wherein each of the images has a different modality;
调用由多个分层组成的分层融合模型对各所述图像分块组进行识别处理，其中，一个分层用于对应识别一个图像分块组，且各所述分层具有预设识别顺序，任意一个所述分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组；Invoke a layered fusion model composed of multiple layers to perform identification processing on each of the image block groups, wherein one layer is used to identify one image block group correspondingly, and each layer has a preset identification sequence , after any one of the image block groups corresponding to the hierarchical recognition and outputting the hierarchical recognition result, the layer next to the sequence begins to identify the corresponding image block group;
在确定满足所述分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。When it is determined that the condition for the hierarchical fusion model to stop performing the recognition processing is satisfied, the biometric recognition result is determined based on the currently obtained hierarchical recognition result.
第二方面，本发明提供了一种生物特征识别装置，该装置包括：In a second aspect, the present invention provides a biometric identification device, the device comprising:
分块单元，用于对目标生物特征的至少两个图像进行分块处理，形成各所述图像各自的图像分块组，其中，各所述图像具有不同的模态；a block unit, configured to perform block processing on at least two images of the target biometric feature to form respective image block groups for each of the images, wherein each of the images has different modalities;
调用单元，用于调用由多个分层组成的分层融合模型对各所述图像分块组进行识别处理，其中，一个分层用于对应识别一个图像分块组，且各所述分层具有预设识别顺序，任意一个所述分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组；a calling unit, configured to call a layered fusion model composed of multiple layers to perform identification processing on each of the image block groups, wherein one layer is used to identify one image block group correspondingly, and each layer Having a preset recognition sequence, after any one of the layered recognition corresponding image block groups and outputting the layered recognition result, the layer next to the sequence starts to recognize the corresponding image block group;
确定单元，用于在确定满足所述分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。The determining unit is configured to determine the biometric identification result based on the currently obtained hierarchical identification result when it is determined that the condition for stopping the identification processing of the layered fusion model is satisfied.
第三方面，本发明提供了一种计算机可读存储介质，所述存储介质包括存储的程序，其中，在所述程序运行时控制所述存储介质所在设备执行第一方面所述的生物特征识别方法。In a third aspect, the present invention provides a computer-readable storage medium, the storage medium includes a stored program, wherein when the program runs, the device where the storage medium is located is controlled to perform the biometric identification described in the first aspect method.
第四方面，本发明提供了一种电子设备，所述电子设备包括：In a fourth aspect, the present invention provides an electronic device, the electronic device comprising:
存储器，用于存储程序；memory for storing programs;
处理器，耦合至所述存储器，用于运行所述程序以执行第一方面所述的生物特征识别方法。A processor, coupled to the memory, is configured to run the program to execute the biometric identification method of the first aspect.
借由上述技术方案，本发明提供的生物特征识别方法及装置，在需要进行生物特征识别时，对目标生物特征的多个不同模态的图像进行分块处理形成各图像各自的图像分块组。然后调用由多个分层组成的分层融合模型对各图像分块组进行识别处理。分层融合模型中的一个分层用于对应识别一个图像分块组，且各分层具有预设识别顺序，在分层融合模型进行生物特征识别时，任意一个分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层才开始识别对应的图像分块组。在确定满足分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。可见，本发明为了减少光照、生物特征采集角度等环境因素对生物特征识别准确度的影响，对目标生物特征的多个不同模态的图像的分块处理后的图像分块组进行生物特征识别。另外，为了提高生物特征识别的速度，一旦当前已得到的所有分层识别结果满足目标生物特征的生物特征识别结果的生成要求，则即可停止分层融合模型的生物特征识别，无需分层融合模型对所有图像分块组均识别完成，便可基于当前已得到的分层识别结果，确定生物特征识别结果，由此减少生物特征识别的识别时间，加快生物特征识别的速度。综上，本发明提供的方案能够在提高生物特征识别准确度的同时，提高生物特征识别的识别速度。With the above technical solutions, the biometric identification method and device provided by the present invention, when biometric identification is required, perform block processing on multiple images of different modalities of the target biometric to form respective image block groups for each image. . Then, a hierarchical fusion model composed of multiple layers is called to identify each image block group. One layer in the layered fusion model is used to identify an image block group correspondingly, and each layer has a preset recognition order. When the layered fusion model performs biometric recognition, any layer identifies the corresponding image block. After grouping and outputting the layer recognition result, the layer next to it in sequence starts to recognize the corresponding image block group. When it is determined that the condition for the hierarchical fusion model to stop performing the recognition processing is satisfied, the biometric recognition result is determined based on the currently obtained hierarchical recognition result. It can be seen that, in order to reduce the influence of environmental factors such as illumination and biometric acquisition angle on the accuracy of biometric identification, the present invention performs biometric identification on the image block groups after the block processing of multiple images of different modalities of the target biometric. . In addition, in order to improve the speed of biometric recognition, once all the currently obtained hierarchical recognition results meet the generation requirements of the biometric recognition results of the target biometric, the biometric recognition of the hierarchical fusion model can be stopped without hierarchical fusion. After the model has identified all image block groups, the biometric identification results can be determined based on the currently obtained hierarchical identification results, thereby reducing the identification time of biometric identification and accelerating the speed of biometric identification. To sum up, the solution provided by the present invention can improve the recognition speed of biometric identification while improving the accuracy of biometric identification.
上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，而可依照说明书的内容予以实施，并且为了让本发明的上述和其它目的、特征和优点能够更明显易懂，以下特举本发明的具体实施方式。The above description is only an overview of the technical solutions of the present invention. In order to understand the technical means of the present invention more clearly, it can be implemented according to the content of the description, and in order to make the above and other objects, features and advantages of the present invention more obvious and easy to understand , the following specific embodiments of the present invention are given.
附图说明Description of drawings
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图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。In order to illustrate the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are For some embodiments of the present invention, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.
图1示出了本发明一个实施例提供的一种生物特征识别方法的流程图；FIG. 1 shows a flowchart of a biometric identification method provided by an embodiment of the present invention;
图2示出了本发明另一个实施例提供的一种生物特征识别方法的流程图；2 shows a flowchart of a biometric identification method provided by another embodiment of the present invention;
图3示出了本发明一个实施例提供的一种多光谱掌纹采集设备的示意图；3 shows a schematic diagram of a multispectral palmprint collection device provided by an embodiment of the present invention;
图4示出了本发明一个实施例提供的一种生物特征识别装置的结构示意图；4 shows a schematic structural diagram of a biometric identification device provided by an embodiment of the present invention;
图5示出了本发明另一个实施例提供的一种生物特征识别装置的结构示意图。FIG. 5 shows a schematic structural diagram of a biometric identification device provided by another embodiment of the present invention.
具体实施方式Detailed ways
下面将参照附图更加详细地描述本公开的示例性实施例。虽然附图中显示了本公开的示例性实施例，然而应当理解，可以以各种形式实现本公开而不应被这里阐述的实施例所限制。相反，提供这些实施例是为了能够更透彻地理解本公开，并且能够将本公开的范围完整的传达给本领域的技术人员。Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure will be more thoroughly understood, and will fully convey the scope of the present disclosure to those skilled in the art.
人体存在诸如指纹、掌纹、虹膜等生物特征。人体的生物特征是人体所固有的生理特征，具有唯一性、隐私性、不可更改性等特性，因此在身份识别领域被广泛应用。The human body has biological features such as fingerprints, palm prints, and iris. The biological characteristics of the human body are inherent physiological characteristics of the human body, and have the characteristics of uniqueness, privacy, and immutability, so they are widely used in the field of identity recognition.
目前，在对人体的生物特征进行生物特征识别时，通常采用的生物特征识别方法包括如下两种：第一种，使用生物特征识别模型对生物特征的单张图像进行生物特征识别，此种方式，虽然生物特征识别速度较快，但是由于图像采集可能受到诸如光照、生物特征采集角度、旋转等因素的影响，造成图像的质量较差，从而导致生物特征识别的准确度较低。第二种，对单张图像进行分块，然后使用生物特征识别模型对图像分块进行生物特征识别。此种方式虽然生物特征识别相对于第一种方法的准确度有所提高，但是由于图像分块的数量众多，严重影响了生物特征识别的速度，导致生物特征识别的速度较低。At present, when the biometrics of the human body are biometrically identified, the biometric identification methods usually used include the following two: First, the biometric identification model is used to perform biometric identification on a single image of the biometric. This method , Although the biometric recognition speed is fast, because the image acquisition may be affected by factors such as illumination, biometric acquisition angle, rotation, etc., the quality of the image is poor, resulting in low biometric recognition accuracy. The second is to block a single image, and then use the biometric recognition model to perform biometric identification of the image blocks. Although the accuracy of biometric identification in this method is improved compared to the first method, due to the large number of image blocks, the speed of biometric identification is seriously affected, resulting in a low speed of biometric identification.
可见，现有的生物特征识别方法存在生物特征识别准确度低和生物特征识别速度低的缺陷。为了克服上述缺陷，本发明实施例提出了一种生物特征识别方法及装置，以在提高生物特征识别准确度的同时，提高生物特征识别的识别速度。本发明实施例提出的生物特征识别方法及装置的具体应用场景可以基于业务要求确定，本实施例中不做具体限定。示例性的，本发明实施例提出的生物特征识别方法及装置可应为于银行、军事等身份识别场景中。下面对本发明实施例提出的生物特征识别方法及装置进行具体说明。It can be seen that the existing biometric identification methods have the defects of low biometric identification accuracy and low biometric identification speed. In order to overcome the above-mentioned defects, the embodiments of the present invention provide a method and device for biometric identification, so as to improve the identification speed of biometric identification while improving the accuracy of biometric identification. The specific application scenarios of the biometric identification method and apparatus proposed in the embodiments of the present invention may be determined based on business requirements, which are not specifically limited in this embodiment. Exemplarily, the biometric identification method and device provided by the embodiments of the present invention may be used in identification scenarios such as banks and military. The biometric identification method and device provided by the embodiments of the present invention will be specifically described below.
如图1所示，本发明实施例提供了一种生物特征识别方法，该方法主要包括：As shown in FIG. 1, an embodiment of the present invention provides a biometric identification method, and the method mainly includes:
101、对目标生物特征的至少两个图像进行分块处理，形成各所述图像各自的图像分块组，其中，各所述图像具有不同的模态。101. Perform block processing on at least two images of the target biometric feature to form respective image block groups for each of the images, wherein each of the images has different modalities.
在实际应用中，生物特征识别是进行身份识别的前提，因此目标生物特征为待进行身份识别的人体所具有的生物特征。目标生物特征的选定与具体身份识别场景有关，本实施例不做具体限定。可选的，目标生物特征为如下中的任意一种：指纹、掌纹和虹膜。In practical applications, biometric identification is the premise of identification, so the target biometrics are the biometrics of the human body to be identified. The selection of the target biometric feature is related to a specific identity recognition scenario, which is not specifically limited in this embodiment. Optionally, the target biometric feature is any one of the following: fingerprint, palm print and iris.
示例性的，本发明实施例提供的生物特征识别方法所针对的目标生物特征为掌纹。选取掌纹的优势在于如下几点：一是，手掌的尺寸相对于手指和眼睛来说，其尺寸加大，因此获取到的掌纹图像中可以包括更为丰富的特征。二是，掌纹的获取相对于指纹来说其安全性更为高。指纹所在的手指肚是凸起的，因此获取指纹时需要将手指肚按压在指纹获取设备的指纹获取面上，才可获取到指纹，指纹容易残留在指纹获取面上，存在被恶意盗取的可能。而掌纹所在的掌心面趋近于一个平面，所以掌纹的获取可通过无接触获得，从而避免掌纹残留在掌纹获取机器上被恶意人员盗取。第三，掌纹的获取相对于虹膜来说其获取更为便利。虹膜获取时需要将眼睛对准虹膜获取设备，用户的脑袋位置必然受限，进而造成用户的行动受限，导致用户体验感较差。而掌纹获取时仅需要用户伸出手掌，用户的行动受限性较小，用户体验感相对较好。第四，掌纹具有独一无二的特性。即使人脸相貌完全相同或近似的人，掌纹也是独一无二的，其安全性更高。Exemplarily, the target biometric feature targeted by the biometric identification method provided by the embodiment of the present invention is a palm print. The advantages of selecting palm prints lie in the following points: First, the size of the palm is larger than that of fingers and eyes, so the obtained palm print image can include more abundant features. Second, palmprints are more secure than fingerprints. The belly of the finger where the fingerprint is located is raised, so when acquiring the fingerprint, you need to press the belly of the finger on the fingerprint acquisition surface of the fingerprint acquisition device to obtain the fingerprint. The fingerprint is easy to remain on the fingerprint acquisition surface, and it may be maliciously stolen. possible. The palm surface where the palmprint is located is close to a plane, so the palmprint can be obtained without contact, so as to avoid the palmprint remaining on the palmprint acquisition machine and being stolen by malicious personnel. Third, the acquisition of palm print is more convenient than that of iris. When the iris is acquired, the eyes need to be aligned with the iris acquisition device, and the position of the user's head is bound to be limited, which in turn causes the user's movement to be limited, resulting in a poor user experience. The palmprint acquisition only requires the user to extend the palm, the user's movement is less restricted, and the user experience is relatively good. Fourth, palm prints have unique characteristics. Even if the faces of the people are the same or similar, the palm print is unique, and its security is higher.
示例性的，身份识别场景为门禁系统身份识别，掌纹因其具有丰富的特征和易获取的特性，将其作为门禁系统身份识别所需的目标生物特征。Exemplarily, the identification scenario is the identification of the access control system, and the palm print is used as the target biometric feature required for the identification of the access control system because of its rich features and easy acquisition characteristics.
在确定目标生物特征之后，需要获取待进行身份识别的人体的目标生成特征的图像。为了减少光照、生物特征采集角度等环境因素对生物特征识别准确度的影响，获取的目标生物特征的图像的数量为两个或两个以上，且各图像具有不同的模态。其中，模态限定了图像的采集场景，也就是各图像的采集场景不同。After the target biometric feature is determined, an image of the target generation feature of the human body to be identified needs to be acquired. In order to reduce the influence of environmental factors such as illumination and biometric acquisition angle on the accuracy of biometric identification, the number of acquired images of target biometrics is two or more, and each image has a different mode. Among them, the modal defines the image collection scene, that is, the collection scene of each image is different.
示例性的，目标生物特征为掌纹，获取了掌纹的三个图像，分别为图像1、图像2和图像3。其中，图像1对应的模态为蓝色光谱模态，也就是说，图像1是在蓝色光谱下采集的图像。图像2对应的模态为绿色光谱模态，也就是说，图像2是在绿色光谱下采集的图像。图像3对应的模态为红色光谱模态，也就是说，图像3是在红色光谱下采集的图像。Exemplarily, the target biological feature is a palm print, and three images of the palm print are acquired, namely image 1, image 2, and image 3. Among them, the mode corresponding to the image 1 is the blue spectrum mode, that is, the image 1 is an image collected under the blue spectrum. The mode corresponding to image 2 is the green spectrum mode, that is, image 2 is an image collected under the green spectrum. The mode corresponding to image 3 is the red spectrum mode, that is, image 3 is an image collected under the red spectrum.
在获取到目标生物特征的图像之后，需要对所获取的图像进行分块处理，形成各图像各自对应的图像分块组。下面对图像分块处理方法进行说明，该方法为：After the image of the target biometric feature is acquired, it is necessary to perform block processing on the acquired image to form an image block group corresponding to each image. The following describes the image block processing method, which is as follows:
针对每一个图像均执行：设定图像的图像分块数量，基于图像分块数量和图像大小确定分块大小，基于分块大小对图像进行分块。示例性的，设定图像1的图像分块数量为64个，图像1的大小为16×16，则将图像1分块处理为64个大小为2×2的图像分块，图像1的图像分块组中包括64个大小为2×2的图像分块。Execute for each image: set the number of image blocks of the image, determine the block size based on the number of image blocks and the image size, and divide the image based on the block size. Exemplarily, if the number of image blocks of image 1 is 64 and the size of image 1 is 16×16, then image 1 is divided into 64 image blocks of size 2×2. The block group includes 64 image blocks of size 2×2.
需要说明的是，目标生物特征的各图像的大小均相同，目标生物特征的各图像的图像分块组中图像分块的数量和大小可存在如下几种：第一种，各图像的图像分块组中图像分块数量均不同，示例性的，存在图像1、图像2和图像3，其中，图像1的图像分块组中包括4个大小为8×8的图像分块，图像2的图像分块组中包括16个大小为4×4的图像分块，图像3的图像分块组中包括64个大小为2×2的图像分块。第二种，各图像的图像分块组中图像分块数量均相同，示例性的，存在图像1、图像2和图像3，图像1、图像2和图像3的图像分块组中均包括64个大小为2×2的图像分块。第三种，各图像的图像分块组中图像分块数量有相同的，也有不同的，示例性的，存在图像1、图像2和图像3，图像1的图像分块组中包括4个大小为8×8的图像分块，图像2和图像3的图像分块组中均包括64个大小为2×2的图像分块。It should be noted that the size of each image of the target biometric feature is the same, and the number and size of the image blocks in the image block group of each image of the target biometric feature can be as follows: The number of image blocks in the block group is different. Exemplarily, there are image 1, image 2 and image 3, wherein the image block group of image 1 includes 4 image blocks with a size of 8 × 8, and the image block of image 2 The image block group includes 16 image blocks with a size of 4×4, and the image block group of image 3 includes 64 image blocks with a size of 2×2. Second, the number of image blocks in the image block group of each image is the same, exemplarily, there are image 1, image 2 and image 3, and the image block groups of image 1, image 2 and image 3 all include 64 an image block of size 2×2. The third type, the number of image blocks in the image block group of each image has the same or different number, for example, there are image 1, image 2 and image 3, and the image block group of image 1 includes 4 sizes It is an 8×8 image block, and the image block groups of image 2 and image 3 both include 64 image blocks with a size of 2×2.
102、调用由多个分层组成的分层融合模型对各所述图像分块组进行识别处理，其中，一个分层用于对应识别一个图像分块组，且各所述分层具有预设识别顺序，任意一个所述分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组。102. Invoke a layered fusion model composed of multiple layers to perform identification processing on each of the image block groups, wherein one layer is used to identify one image block group correspondingly, and each layer has a preset value. Recognition sequence, after any one of the layers recognizes the corresponding image block group and outputs the layer recognition result, the layer next to the sequence starts to recognize the corresponding image block group.
分层融合模型由多个分层组成，其用于对图像进行生物特征识别。分层融合模型中的各分层具有预设识别顺序，一个分层用于对应识别一个图像分块组。在预设识别顺序中，任意一个分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层才开始识别对应的图像分块组。此种分层按照预设识别顺序进行识别方式的目的在于，在分层识别的过程中，一旦某个分层的分层识别结果输出后，使得当前已得到的所有分层识别结果满足目标生物特征的生物特征识别结果的生成要求，则即可停止分层融合模型的生物特征识别，基于当前已得到的分层识别结果，便可确定生物特征识别结果，由此减少生物特征识别的识别时间，加快生物特征识别的速度。The layered fusion model consists of multiple layers, which are used for biometric identification of images. Each layer in the layered fusion model has a preset recognition order, and one layer is used to identify one image block group correspondingly. In the preset recognition sequence, after any layer recognizes the corresponding image block group and outputs the layer recognition result, the layer next to the sequence starts to recognize the corresponding image block group. The purpose of this layered recognition method according to the preset recognition order is that, in the process of layered recognition, once the layered recognition results of a certain layer are output, all the layered recognition results that have been obtained so far meet the target biological requirements. If the biometric identification result of the feature is required to be generated, the biometric identification of the hierarchical fusion model can be stopped, and the biometric identification result can be determined based on the currently obtained hierarchical identification result, thereby reducing the identification time of biometric identification. , to speed up biometric identification.
需要说明的是，分层融合模型的类型可以基于具体业务需求确定，本实施例不做具体限定。可选的，分层融合模型可以为分层式变换(Transformer)模型或多层感知模型(Multilayer Perceptron，MLP)。It should be noted that the type of the layered fusion model may be determined based on specific business requirements, which is not specifically limited in this embodiment. Optionally, the layered fusion model may be a layered transform (Transformer) model or a multi-layer perceptron (Multilayer Perceptron, MLP).
下面对调用由多个分层组成的分层融合模型对各图像分块组进行识别处理的具体过程进行说明，该过程包括如下步骤一至步骤二：The following describes the specific process of invoking a layered fusion model composed of multiple layers to identify and process each image block group. The process includes the following steps 1 to 2:
步骤一，确定分层融合模型中各分层对应的图像分块组。Step 1: Determine the image block group corresponding to each layer in the layered fusion model.
在实际应用中，确定分层融合模型中各分层对应的图像分块组的过程与各图像分块组中图像分块数量有关，因此该确定过程存在如下几种情况：In practical applications, the process of determining the image block group corresponding to each layer in the layered fusion model is related to the number of image blocks in each image block group. Therefore, the determination process has the following situations:
第一种，在各图像分块组中的图像分块数量不同时，则在确定各分层对应的图像分块组时，在各分层的预设识别顺序中排序越靠前的分层，其对应的图像分块组中的图像分块数量越少。First, when the number of image blocks in each image block group is different, when the image block group corresponding to each layer is determined, the layer that is ranked higher in the preset recognition order of each layer is sorted. , the smaller the number of image blocks in the corresponding image block group.
由于预设识别顺序中排序越靠前的分层，其越先进行生物特征识别，在分层识别的过程中，一旦某个分层的分层识别结果输出后，使得当前已得到的所有分层识别结果满足目标生物特征的生物特征识别结果的生成要求，则即可停止顺序位于其后的分层的生物特征识别，基于当前已得到的分层识别结果，确定生物特征识别结果。由此减少生物特征识别的识别时间，加快生物特征识别的速度。Since the higher-ranked layer in the preset recognition order, the biometric recognition is performed first. If the layer recognition result satisfies the generation requirements of the biometric recognition result of the target biometric, the biometric recognition of the subsequent layers can be stopped, and the biometric recognition result is determined based on the currently obtained layered recognition result. Thereby, the identification time of biometric identification is reduced, and the speed of biometric identification is accelerated.
示例性的，分层融合模型由分层1、分层2和分层3组成，预设识别顺序中各分层识别先后排序为：分层1、分层2和分层3。存在图像1、图像2和图像3，各图像的图像分块组中图像分块数量从少到多的顺序为：图像1、图像2和图像3。则确定分层1对应图像1的图像分块组、分层2对应图像2的图像分块组和分层3对应图像3的图像分块组。Exemplarily, the layered fusion model is composed of layer 1, layer 2, and layer 3, and the recognition sequence of each layer in the preset recognition sequence is: layer 1, layer 2, and layer 3. There are image 1 , image 2 and image 3 , and the order of the number of image blocks in the image block group of each image is: image 1 , image 2 and image 3 . Then it is determined that layer 1 corresponds to the image block group of image 1, layer 2 corresponds to the image block group of image 2, and layer 3 corresponds to the image block group of image 3.
第二种，在各图像分块组中的图像分块数量相同时，确定各分层对应的图像分块组的方法包括如下两种：一种是，在确定各分层对应的图像分块组时，随机确定各分层对应图像分块组，在确定时保证各分层与图像分块组之间的对应关系唯一即可。另一种是，根据图像分块组所属图像的模态确定各分层的图像分块组，比如，模态之间具有设定的模态识别顺序，则基于模态识别顺序为各分层确定对应的图像分块组。示例性，图像1、图像2和图像3。其中，图像1对应的模态为蓝色光谱模态，图像2对应的模态为绿色光谱模态，图像3对应的模态为红色光谱模态。模态识别顺序从前到后为：蓝色光谱模态、绿色光谱模态、红色光谱模态，预设识别顺序中各分层识别先后排序为：分层1、分层2和分层3。则确定分层1对应图像1的图像分块组、分层2对应图像2的图像分块组和分层3对应图像3的图像分块组。Second, when the number of image blocks in each image block group is the same, the method for determining the image block group corresponding to each layer includes the following two methods: one is, when determining the image block corresponding to each layer When grouping, the image block group corresponding to each layer is randomly determined, and the corresponding relationship between each layer and the image block group can be guaranteed to be unique when determining. The other is to determine the image block group of each layer according to the modality of the image to which the image block group belongs. Determine the corresponding image tile group. Exemplarily, Image 1, Image 2, and Image 3. The mode corresponding to image 1 is the blue spectral mode, the mode corresponding to image 2 is the green spectral mode, and the mode corresponding to image 3 is the red spectral mode. The modal recognition sequence from front to back is: blue spectral modal, green spectral modal, and red spectral modal. Then it is determined that layer 1 corresponds to the image block group of image 1, layer 2 corresponds to the image block group of image 2, and layer 3 corresponds to the image block group of image 3.
第三种，各图像的图像分块组中图像分块数量有相同的，也有不同的。存在图像1、图像2和图像3，图像1的图像分块组中包括4个图像分块，图像2和图像3的图像分块组中均包括64个图像分块。预设识别顺序中各分层识别先后排序为：分层1、分层2和分层3。各图像的图像分块组中图像分块数量从少到多的顺序为：图像1、图像2和图像3，此排序中图像2和图像3分块数量相同，随机确定图像2排在图像3之前。则确定分层1对应图像1的图像分块组、分层2对应图像2的图像分块组和分层3对应图像3的图像分块组。Third, the number of image blocks in the image block group of each image may be the same or different. There are image 1, image 2 and image 3, the image block group of image 1 includes 4 image blocks, and the image block groups of image 2 and image 3 both include 64 image blocks. The recognition sequence of each layer in the preset recognition sequence is: layer 1, layer 2, and layer 3. The order of the number of image blocks in the image block group of each image is: image 1, image 2 and image 3. In this order, the number of image blocks of image 2 and image 3 is the same, and image 2 is randomly determined to be ranked in image 3. Before. Then it is determined that layer 1 corresponds to the image block group of image 1, layer 2 corresponds to the image block group of image 2, and layer 3 corresponds to the image block group of image 3.
步骤二，依次调用各分层对应识别其各自对应的图像分块组。In step 2, each layer is called in turn to identify its corresponding image block group.
在确定各分层对应的图像分块组之后，依次调用各分层对应识别其各自对应的图像分块组，需要强调的是，在调用各分层时，当前使用的分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组。After determining the image block group corresponding to each layer, call each layer in turn to identify its corresponding image block group. It should be emphasized that when calling each layer, the currently used layer identifies the corresponding image. After the block group is divided and the layer recognition result is output, the layer next to it in sequence starts to recognize the corresponding image block group.
103、在确定满足所述分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。103. When it is determined that the condition for stopping the identification processing by the layered fusion model is satisfied, determine the biometric identification result based on the currently obtained layered identification result.
为了及时确定生物特征识别结果，则需要确定当前是否满足分层融合模型停止进行识别处理的条件，以在确定满足分层融合模型停止进行识别处理的条件时，停止调用分层融合模型进行生物特征识别。确定满足分层融合模型停止进行识别处理的条件的过程包括如下步骤一至步骤五：In order to determine the biometric identification result in time, it is necessary to determine whether the current condition for the hierarchical fusion model to stop performing identification processing is met, so that when it is determined that the condition for the hierarchical fusion model to stop performing identification processing is met, stop calling the hierarchical fusion model for biometric processing. identify. The process of determining that the condition for the hierarchical fusion model to stop performing the identification processing is satisfied includes the following steps 1 to 5:
步骤一，确定当前最新得到的分层识别结果。Step 1: Determine the latest obtained hierarchical identification result.
当前最新得到的分层识别结果为最新进行生物特征识别的分层所输出的分层识别结果。The current latest obtained hierarchical identification result is the hierarchical identification result output by the latest biometric identification layer.
分层识别结果中包括有识别分值和目标对象，识别分值用于体现分层识别结果对应的图像为目标对象的生物特征的图像的概率。示例性的，分层1对图像1的分层识别结果中包括识别分值“80％”和目标对象“张三”，说明图像1为张三的生物特征的图像的概率为80％。The hierarchical recognition result includes a recognition score and a target object, and the recognition score is used to reflect the probability that the image corresponding to the hierarchical recognition result is an image of the biological feature of the target object. Exemplarily, the hierarchical recognition result of image 1 by layer 1 includes the recognition score "80%" and the target object "Zhang San", indicating that the probability of image 1 being an image of Zhang San's biometric features is 80%.
步骤二，判断当前最新得到的分层识别结果包括的识别分值是否达到第一阈值，若未达到，执行步骤三；否则，执行步骤四。In step 2, it is judged whether the recognition score included in the newly obtained hierarchical recognition result reaches the first threshold, if not, step 3 is performed; otherwise, step 4 is performed.
分层融合模型由多个分层组成，各分层输出的分层识别结果中的识别分值的总和理论上来说应为一个设定数值，比如100％。因此，需要设定第一阈值，若一个分层的分层识别结果中的识别分值达到第一阈值，则说明该分层识别结果的准确度较高，预设识别顺序中顺序位于其后的分层即使进行生物特征识别，得到的分层识别结果的准确度也不会高于其输出的分层识别结果的准确度。因此，为了及时确定生物特征识别结果，需要判断当前最新得到的分层识别结果包括的识别分值是否达到第一阈值。The layered fusion model is composed of multiple layers, and the sum of the recognition scores in the layered recognition results output by each layer should theoretically be a set value, such as 100%. Therefore, it is necessary to set a first threshold. If the recognition score in a hierarchical recognition result of a hierarchical layer reaches the first threshold, it means that the accuracy of the hierarchical recognition result is high, and the sequence of the predetermined recognition sequence is next. Even if biometric identification is performed, the accuracy of the obtained layered identification results will not be higher than the accuracy of the output layered identification results. Therefore, in order to determine the biometric identification result in time, it is necessary to determine whether the identification score included in the newly obtained hierarchical identification result reaches the first threshold.
在判断出当前最新得到的分层识别结果包括的识别分值达到第一阈值，说明该分层识别结果的准确度较高，可直接将该分层识别结果确定为目标生物特征的生物特征识别结果，分层融合模型可停止进行识别处理，因此执行步骤四。When it is judged that the recognition score included in the newly obtained hierarchical recognition result reaches the first threshold, it means that the accuracy of the hierarchical recognition result is relatively high, and the hierarchical recognition result can be directly determined as the biometric identification of the target biometrics. As a result, the hierarchical fusion model can stop the recognition process, so step four is performed.
在判断出当前最新得到的分层识别结果包括的识别分值未达到第一阈值，说明尚不能基于当前最新得到的分层识别结果确定目标生物特征的生物特征识别结果，因此需要执行步骤三。When it is determined that the recognition score included in the newly obtained hierarchical recognition result does not reach the first threshold, it means that the biometric recognition result of the target biological feature cannot be determined based on the currently newly obtained hierarchical recognition result, so step 3 needs to be performed.
示例性的，分层融合模型由三个分层组成，预设识别顺序中各分层识别先后排序为：分层1、分层2和分层3。原则上各分层输出的分层识别结果中的识别分值的总和理论上来说应为100％。当前最新得到的分层识别结果为分层1输出的分层识别结果1。经判断，当前最新得到的分层识别结果包括的识别分值“80％”达到第一阈值“60％”，说明分层识别结果1的准确度较高，可直接将分层识别结果1确定为目标生物特征的生物特征识别结果。Exemplarily, the layered fusion model consists of three layers, and the recognition sequence of each layer in the preset recognition sequence is: layer 1, layer 2, and layer 3. In principle, the sum of the recognition scores in the hierarchical recognition results output by each hierarchical layer should theoretically be 100%. The latest layer recognition result obtained at present is the layer recognition result 1 output by layer 1. It is judged that the recognition score "80%" included in the latest hierarchical recognition result has reached the first threshold of "60%", indicating that the accuracy of hierarchical recognition result 1 is relatively high, and the hierarchical recognition result 1 can be directly determined. The biometric identification result for the target biometric.
步骤三，判断当前已得到的分层识别结果中是否存在至少两个目标分层识别结果，其中，所有目标分层识别结果中的识别对象相同，且所有目标分层识别结果中的识别分值均达到第二阈值，第二阈值小于第一阈值。若存在，执行步骤四；否则，执行步骤五。Step 3: Judge whether there are at least two target hierarchical recognition results in the currently obtained hierarchical recognition results, wherein the recognition objects in all target hierarchical recognition results are the same, and the recognition scores in all target hierarchical recognition results are the same. Both reach the second threshold, and the second threshold is smaller than the first threshold. If it exists, go to step four; otherwise, go to step five.
在判断出当前最新得到的分层识别结果包括的识别分值未达到第一阈值时，说明不能直接依据当前最新得到的分层识别结果确定目标生物特征的生物特征识别结果。因此，需要判断当前已得到的分层识别结果中是否存在至少两个目标分层识别结果。When it is judged that the recognition score included in the currently newly obtained hierarchical recognition result does not reach the first threshold, it means that the biometric recognition result of the target biological feature cannot be directly determined according to the currently newly obtained hierarchical recognition result. Therefore, it is necessary to judge whether there are at least two target hierarchical recognition results in the currently obtained hierarchical recognition results.
这里所述的所有目标分层识别结果中的识别对象相同，且所有目标分层识别结果中的识别分值均达到第二阈值，第二阈值小于第一阈值。设置第二阈值的原则是，保证所有目标分层识别结果中的识别分值取第二阈值时，目标分层识别结果中的识别分值的总和大于各分层输出的分层识别结果中的识别分值的总和的一半。示例性，分层融合模型中的分层数量为4个，第二阈值为30％，也就是说，当前已得到的分层识别结果中存在两个识别分值达到30％的分层识别结果时，则确定满足分层融合模型停止进行识别处理的条件。The recognition objects in all the target hierarchical recognition results described here are the same, and the recognition scores in all target hierarchical recognition results reach the second threshold, and the second threshold is smaller than the first threshold. The principle of setting the second threshold is to ensure that when the recognition scores in all the target hierarchical recognition results take the second threshold, the sum of the recognition scores in the target hierarchical recognition results is greater than that in the hierarchical recognition results output by each layer. Identify half the sum of the points. Exemplarily, the number of layers in the layered fusion model is 4, and the second threshold is 30%, that is, there are two layered recognition results whose recognition scores reach 30% in the currently obtained layered recognition results. When , it is determined that the condition for the hierarchical fusion model to stop performing the identification process is satisfied.
在判断出当前已得到的分层识别结果中存在至少两个目标分层识别结果，说明可依据当前已得到的分层识别结果确定目标生物特征的生物特征识别结果，分层融合模型可停止进行识别处理，因此执行步骤四。It is judged that there are at least two target hierarchical recognition results in the currently obtained hierarchical recognition results, indicating that the biometric recognition results of the target biometrics can be determined according to the currently obtained hierarchical recognition results, and the hierarchical fusion model can be stopped. The identification process is performed, so step four is performed.
在判断出当前已得到的分层识别结果中不存在至少两个目标分层识别结果，说明尚不能依据当前已得到的分层识别结果确定目标生物特征的生物特征识别结果，因此需要执行步骤五。It is judged that there are no at least two target hierarchical recognition results in the currently obtained hierarchical recognition results, indicating that the biometric recognition results of the target biometrics cannot be determined according to the currently obtained hierarchical recognition results, so step 5 needs to be performed. .
步骤四，确定满足分层融合模型停止进行识别处理的条件，并停止调用分层融合模型，结束当前流程。Step 4: It is determined that the condition for stopping the identification processing of the layered fusion model is satisfied, and the calling of the layered fusion model is stopped, and the current process is ended.
确定满足分层融合模型停止进行识别处理的条件，则可停止调用分层融合模型，从而提高分层融合模型的生物特征识别的速度。If it is determined that the condition for the hierarchical fusion model to stop performing identification processing is satisfied, the hierarchical fusion model can be stopped, thereby improving the speed of biometric identification of the hierarchical fusion model.
步骤五，判断当前最新得到的分层识别结果是否为分层融合模型中预设识别顺序中最后一位的分层输出的分层识别结果；若是，结束当前流程；否则，执行步骤一。Step 5: Determine whether the currently newly obtained layered identification result is the layered identification result of the last layer output in the preset identification sequence in the layered fusion model; if so, end the current process; otherwise, go to step one.
在判断出当前最新得到的分层识别结果为分层融合模型中预设识别顺序中最后一位的分层输出的分层识别结果，则说明分层融合模型中所有的分层均完成的生物特征识别，需要根据所有分层的分层识别结果，确定生物特征识别结果。When it is judged that the newly obtained layered recognition result is the layered recognition result of the last layer output in the preset recognition sequence in the layered fusion model, it means that all layers in the layered fusion model are completed. For feature identification, the biometric identification results need to be determined according to the hierarchical identification results of all layers.
在判断出当前最新得到的分层识别结果不是分层融合模型中预设识别顺序中最后一位的分层输出的分层识别结果，则继续调用分层融合模型进行生物特征识别，以使顺序紧邻当前输出分层识别结果的分层之后的分层继续识别对应的图像分块组。After judging that the latest obtained hierarchical recognition result is not the hierarchical recognition result of the last hierarchical output in the preset recognition sequence in the hierarchical fusion model, continue to call the hierarchical fusion model for biometric recognition, so that the sequence The layer immediately following the layer that currently outputs the layer recognition result continues to identify the corresponding image block group.
在经过上述步骤一至步骤五之后，并确定满足分层融合模型停止进行识别处理的条件时，需要基于当前已得到的分层识别结果确定生物特征识别结果，下面对确定生物特征识别结果的具体方法进行说明，该确定方法包括如下几种：After the above steps 1 to 5, and it is determined that the condition for the hierarchical fusion model to stop performing the recognition processing is satisfied, the biometric recognition result needs to be determined based on the currently obtained hierarchical recognition result. The method is described, and the determination method includes the following:
第一种，在判断出当前最新得到的分层识别结果包括的识别分值达到第一阈值时，将当前最新得到的分层识别结果确定为生物特征识别结果。First, when it is determined that the recognition score included in the currently newly obtained hierarchical recognition result reaches the first threshold, the currently newly obtained hierarchical recognition result is determined as the biometric recognition result.
在判断出当前最新得到的分层识别结果包括的识别分值达到第一阈值时，说明当前最新得到的分层识别结果是识别准确度最高的分层识别结果，即使预设识别顺序中顺序位于其后的分层进行生物特征识别，得到的分层识别结果的准确度也不会高于当前最新得到的分层识别结果的准确度，因此将当前最新得到的分层识别结果确定为生物特征识别结果。When it is judged that the recognition score included in the newly obtained hierarchical recognition result reaches the first threshold, it means that the currently newly obtained hierarchical recognition result is the hierarchical recognition result with the highest recognition accuracy, even if the sequence in the preset recognition sequence is located in Biometric identification is performed on subsequent layers, and the accuracy of the obtained layered identification results will not be higher than the accuracy of the current latest layered identification results. Therefore, the current latest layered identification results are determined as biometrics. Identify the results.
第二种，在判断出当前最新得到的分层识别结果包括的识别分值未达到第一阈值，但判断出当前已得到的分层识别结果中存在至少两个目标分层识别结果时，将至少两个目标分层识别结果中识别分值最高的目标分层识别结果，确定为生物特征识别结果。Second, when it is judged that the recognition score included in the newly obtained hierarchical recognition result does not reach the first threshold, but it is judged that there are at least two target hierarchical recognition results in the currently obtained hierarchical recognition results, The target hierarchical recognition result with the highest recognition score among the at least two target hierarchical recognition results is determined as the biometric recognition result.
所有目标分层识别结果中的识别对象相同，原则上各目标分层识别结果均可作为生物特征识别结果。但是为了保证输出生物特征识别结果的准确性，将所有目标分层识别结果中识别分值最高的目标分层识别结果，确定为目标生物特征的生物特征识别结果。The recognition objects in all target hierarchical recognition results are the same. In principle, each target hierarchical recognition result can be used as the biometric recognition result. However, in order to ensure the accuracy of the output biometric identification results, the target hierarchical identification result with the highest recognition score among all the target hierarchical identification results is determined as the biometric identification result of the target biometric.
第三种，在分层融合模型中所有分层均已输出分层识别结果，将当前已得到的分层识别结果中识别分值最高的分层识别结果，确定为生物特征识别结果。The third is that all layers in the layered fusion model have output layered recognition results, and the layered recognition result with the highest recognition score among the currently obtained layered recognition results is determined as the biometric recognition result.
识别分值用于体现分层识别结果对应的图像为目标对象的生物特征的图像的概率。因此，识别分值越高分层识别结果对应的图像为目标对象的生物特征的图像的概率越高，因此，将当前已得到的分层识别结果中识别分值最高的分层识别结果，确定为生物特征识别结果。The recognition score is used to reflect the probability that the image corresponding to the hierarchical recognition result is an image of the biological feature of the target object. Therefore, the higher the recognition score, the higher the probability that the image corresponding to the hierarchical recognition result is an image of the target object's biometrics. Therefore, the hierarchical recognition result with the highest recognition score among the currently obtained hierarchical recognition results is determined. for biometric identification results.
第四种，在分层融合模型中所有分层均已输出分层识别结果，且当前已得到分层识别结果中的识别分值之间的差值均在预设差值范围内，将具有相同识别对象的分层识别结果分为一组，将具有分层识别结果最多的一组中识别分值最高的分层识别结果，确定为生物特征识别结果。Fourth, in the layered fusion model, all layers have output layered recognition results, and the difference between the recognition scores in the currently obtained layered recognition results is within the preset difference range, which will have The hierarchical recognition results of the same recognition object are divided into one group, and the hierarchical recognition result with the highest recognition score in the group with the most hierarchical recognition results is determined as the biometric recognition result.
示例性的，分层融合模型中分层1的分层识别结果1为“张三，24％”，分层2的分层识别结果2为“张三，25％”，分层3的分层识别结果3为“李四，25％”，分层4的分层识别结果4为“王五，26％”，可见，当前已得到分层识别结果中的识别分值之间的差值均在预设差值范围(-3％，3％)内，将具有相同识别对象的分层识别结果分为一组“分层识别结果1和分层识别结果2分为组1，分层识别结果3分为组2，分层识别结果4分为组3”，将具有分层识别结果最多的一组“组1”中识别分值最高的分层识别结果“分层识别结果2”，确定为生物特征识别结果。Exemplarily, in the layered fusion model, the layer recognition result 1 of layer 1 is "Zhang San, 24%", the layer recognition result 2 of layer 2 is "Zhang San, 25%", and the score of layer 3 is "Zhang San, 25%". The layer recognition result 3 is "Li Si, 25%", and the layer recognition result 4 of layer 4 is "Wang Wu, 26%". It can be seen that the difference between the recognition scores in the layer recognition results has been obtained. All within the preset difference range (-3%, 3%), the hierarchical recognition results with the same recognition object are divided into one group. The recognition result 3 is divided into group 2, the hierarchical recognition result 4 is divided into group 3", and the hierarchical recognition result with the highest recognition score in the group "group 1" with the most hierarchical recognition results is divided into "hierarchical recognition result 2" , determined as the biometric identification result.
本发明实施例提供的生物特征识别方法，在需要进行生物特征识别时，对目标生物特征的多个不同模态的图像进行分块处理形成各图像各自的图像分块组。然后调用由多个分层组成的分层融合模型对各图像分块组进行识别处理。分层融合模型中的一个分层用于对应识别一个图像分块组，且各分层具有预设识别顺序，在分层融合模型进行生物特征识别时，任意一个分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层才开始识别对应的图像分块组。在确定满足分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。可见，本发明实施例为了减少光照、生物特征采集角度等环境因素对生物特征识别准确度的影响，对目标生物特征的多个不同模态的图像的分块处理后的图像分块组进行生物特征识别。另外，为了提高生物特征识别的速度，一旦当前已得到的所有分层识别结果满足目标生物特征的生物特征识别结果的生成要求，则即可停止分层融合模型的生物特征识别，无需分层融合模型对所有图像分块组均识别完成，便可基于当前已得到的分层识别结果，确定生物特征识别结果，由此减少生物特征识别的识别时间，加快生物特征识别的速度。综上，本发明实施例提供的方案能够在提高生物特征识别准确度的同时，提高生物特征识别的识别速度。In the biometric identification method provided by the embodiment of the present invention, when biometric identification is required, a plurality of images of different modalities of the target biometric are processed into blocks to form respective image block groups for each image. Then, a hierarchical fusion model composed of multiple layers is called to identify each image block group. One layer in the layered fusion model is used to identify an image block group correspondingly, and each layer has a preset recognition order. When the layered fusion model performs biometric recognition, any layer identifies the corresponding image block. After grouping and outputting the layer recognition result, the layer next to it in sequence starts to recognize the corresponding image block group. When it is determined that the condition for the hierarchical fusion model to stop performing the recognition processing is satisfied, the biometric recognition result is determined based on the currently obtained hierarchical recognition result. It can be seen that in order to reduce the influence of environmental factors such as illumination and biometric acquisition angle on the accuracy of biometric identification, the embodiment of the present invention performs biometric processing on the image block groups after the block processing of images of multiple different modalities of the target biometric. Feature recognition. In addition, in order to improve the speed of biometric recognition, once all the currently obtained hierarchical recognition results meet the generation requirements of the biometric recognition results of the target biometric, the biometric recognition of the hierarchical fusion model can be stopped without hierarchical fusion. After the model has identified all image block groups, the biometric identification results can be determined based on the currently obtained hierarchical identification results, thereby reducing the identification time of biometric identification and accelerating the speed of biometric identification. To sum up, the solutions provided by the embodiments of the present invention can improve the recognition speed of biometrics while improving the accuracy of biometrics.
进一步的，根据图1所示的方法，本发明的另一个实施例还提供了一种生物特征识别方法，如图2所示，该方法主要包括：Further, according to the method shown in FIG. 1, another embodiment of the present invention also provides a biometric identification method, as shown in FIG. 2, the method mainly includes:
201、采集目标生物特征的至少两个图像，其中，各图像具有不同的模态。201. Collect at least two images of the target biological feature, wherein each image has a different modality.
在实际应用中，通过多光谱设备采集目标生物特征的至少两个图像，其中，各图像对应不同的光谱。In practical applications, at least two images of the target biological feature are collected by a multispectral device, wherein each image corresponds to a different spectrum.
示例性的，掌纹由于具有特征丰富、易获取、独一无二的特征，因此目标生物特征选用为掌纹，通过图3所示的多光谱掌纹采集设备采集掌纹。图3中A1为光谱控制元件，A2为多光谱灯，A3为图像采集设备、B为人体手掌放置位置。采集掌纹的图像时，手掌放置时掌纹需要与图像采集设备相对。图像采集设备每采集掌纹的一个图像后，光谱控制元件控制多光谱灯改变一次光谱。经过多光谱设备采集掌纹，获取到了掌纹的三个图像，分别为图像1、图像2和图像3。其中，图像1对应的模态为蓝色光谱模态，也就是说，图像1是在蓝色光谱下采集的图像。图像2对应的模态为绿色光谱模态，也就是说，图像2是在绿色光谱下采集的图像。图像3对应的模态为红色光谱模态，也就是说，图像3是在红色光谱下采集的图像。Exemplarily, the palm print is rich in features, easy to obtain, and unique, so the target biological feature is selected as the palm print, and the palm print is collected by the multi-spectral palm print collection device shown in FIG. 3 . In Figure 3, A1 is the spectral control element, A2 is the multi-spectral lamp, A3 is the image acquisition device, and B is the placement position of the human palm. When capturing images of palm prints, the palm prints need to be opposite to the image capturing device when the palm is placed. After the image acquisition device collects an image of the palmprint, the spectrum control element controls the multi-spectral lamp to change the spectrum once. After collecting palmprints with multispectral equipment, three images of palmprints were obtained, namely image 1, image 2 and image 3. Among them, the mode corresponding to the image 1 is the blue spectrum mode, that is, the image 1 is an image collected under the blue spectrum. The mode corresponding to image 2 is the green spectrum mode, that is, image 2 is an image collected under the green spectrum. The mode corresponding to image 3 is the red spectrum mode, that is, image 3 is an image collected under the red spectrum.
202、对目标生物特征的至少两个图像进行分块处理，形成各图像各自的图像分块组。202. Perform block processing on at least two images of the target biometric feature to form respective image block groups for each image.
203、调用由多个分层组成的分层融合模型对各所述图像分块组进行识别处理，其中，一个分层用于对应识别一个图像分块组，且各分层具有预设识别顺序，任意一个所述分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组。203. Invoke a layered fusion model composed of multiple layers to perform identification processing on each of the image block groups, wherein one layer is used to identify one image block group correspondingly, and each layer has a preset identification sequence , after any one of the layers recognizes the corresponding image block group and outputs the layer recognition result, the layer next in sequence starts to recognize the corresponding image block group.
204、在确定满足所述分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。204. When it is determined that the condition for stopping the identification processing by the layered fusion model is satisfied, determine the biometric identification result based on the currently obtained layered identification result.
上述步骤202至步骤204中的具体说明与图1中步骤101值步骤103的基本相同，因此这里将不再赘述。The specific descriptions of the
进一步的，依据上述方法实施例，本发明的另一个实施例还提供了一种生物特征识别装置，如图4所示，所述装置包括：Further, according to the above method embodiment, another embodiment of the present invention further provides a biometric identification device, as shown in FIG. 4 , the device includes:
分块单元31，用于对目标生物特征的至少两个图像进行分块处理，形成各所述图像各自的图像分块组，其中，各所述图像具有不同的模态；A
调用单元32，用于调用由多个分层组成的分层融合模型对各所述图像分块组进行识别处理，其中，一个分层用于对应识别一个图像分块组，且各所述分层具有预设识别顺序，任意一个所述分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层开始识别对应的图像分块组；The calling
确定单元33，用于在确定满足所述分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。The determining
本发明实施例提供的生物特征识别装置，在需要进行生物特征识别时，对目标生物特征的多个不同模态的图像进行分块处理形成各图像各自的图像分块组。然后调用由多个分层组成的分层融合模型对各图像分块组进行识别处理。分层融合模型中的一个分层用于对应识别一个图像分块组，且各分层具有预设识别顺序，在分层融合模型进行生物特征识别时，任意一个分层识别对应的图像分块组并输出分层识别结果后，顺序紧邻其后的分层才开始识别对应的图像分块组。在确定满足分层融合模型停止进行识别处理的条件时，基于当前已得到的分层识别结果，确定生物特征识别结果。可见，本发明实施例为了减少光照、生物特征采集角度等环境因素对生物特征识别准确度的影响，对目标生物特征的多个不同模态的图像的分块处理后的图像分块组进行生物特征识别。另外，为了提高生物特征识别的速度，一旦当前已得到的所有分层识别结果满足目标生物特征的生物特征识别结果的生成要求，则即可停止分层融合模型的生物特征识别，无需分层融合模型对所有图像分块组均识别完成，便可基于当前已得到的分层识别结果，确定生物特征识别结果，由此减少生物特征识别的识别时间，加快生物特征识别的速度。综上，本发明实施例提供的方案能够在提高生物特征识别准确度的同时，提高生物特征识别的识别速度。The biometric identification device provided by the embodiment of the present invention performs block processing on a plurality of images of different modalities of the target biometric to form respective image block groups for each image when biometric identification is required. Then, a hierarchical fusion model composed of multiple layers is called to identify each image block group. One layer in the layered fusion model is used to identify an image block group correspondingly, and each layer has a preset recognition order. When the layered fusion model performs biometric recognition, any layer identifies the corresponding image block. After grouping and outputting the layer recognition result, the layer next to it in sequence starts to recognize the corresponding image block group. When it is determined that the condition for the hierarchical fusion model to stop performing the recognition processing is satisfied, the biometric recognition result is determined based on the currently obtained hierarchical recognition result. It can be seen that in order to reduce the influence of environmental factors such as illumination and biometric acquisition angle on the accuracy of biometric identification, the embodiment of the present invention performs biometric processing on the image block groups after the block processing of images of multiple different modalities of the target biometric. Feature recognition. In addition, in order to improve the speed of biometric recognition, once all the currently obtained hierarchical recognition results meet the generation requirements of the biometric recognition results of the target biometric, the biometric recognition of the hierarchical fusion model can be stopped without hierarchical fusion. After the model has identified all image block groups, the biometric identification results can be determined based on the currently obtained hierarchical identification results, thereby reducing the identification time of biometric identification and accelerating the speed of biometric identification. To sum up, the solutions provided by the embodiments of the present invention can improve the recognition speed of biometrics while improving the accuracy of biometrics.
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第一确定模块331，用于在判断出当前最新得到的分层识别结果包括的识别分值达到第一阈值时，确定满足所述分层融合模型停止进行识别处理的条件，其中，所述分层识别结果中包括有识别分值和目标对象，所述识别分值用于体现所述分层识别结果对应的图像为所述目标对象的生物特征的图像的概率。The
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第二确定模块332，用于将所述当前最新得到的分层识别结果确定为所述生物特征识别结果。The
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第三确定模块333，用于在判断出当前最新得到的分层识别结果包括的识别分值未达到第一阈值时，判断当前已得到的分层识别结果中是否存在至少两个目标分层识别结果；在判断出当前已得到的分层识别结果中存在所述至少两个目标分层识别结果时，确定满足所述分层融合模型停止进行识别处理的条件。其中，所有目标分层识别结果中的识别对象相同，且所有目标分层识别结果中的识别分值均达到第二阈值，所述第二阈值小于所述第一阈值；The
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第四确定模块334，用于将所述至少两个目标分层识别结果中识别分值最高的目标分层识别结果，确定为所述生物特征识别结果。The
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第五确定模块335，用于在所述分层融合模型中所有分层均已输出分层识别结果，将当前已得到的分层识别结果中识别分值最高的分层识别结果，确定为所述生物特征识别结果。其中，所述分层识别结果中包括有识别分值和目标对象，所述识别分值用于体现所述分层识别结果对应的图像为所述目标对象的生物特征的图像的概率。The
可选的，如图5所示，所述确定单元33包括：Optionally, as shown in FIG. 5 , the determining
第六确定模块336，用于在所述分层融合模型中所有分层均已输出分层识别结果，且当前已得到分层识别结果中的识别分值之间的差值均在预设差值范围内，将具有相同识别对象的分层识别结果分为一组，将具有分层识别结果最多的一组中识别分值最高的分层识别结果，确定为所述生物特征识别结果；其中，所述分层识别结果中包括有识别分值和目标对象，所述识别分值用于体现所述分层识别结果对应的图像为所述目标对象的生物特征的图像的概率。The
可选的，如图5所示，调用单元32，具体用于确定各所述分层对应的图像分块组，其中，在所述预设识别顺序中排序越靠前的分层，其对应的图像分块组中的图像分块数量越少；依次调用各所述分层对应识别其各自对应的图像分块组。Optionally, as shown in FIG. 5 , the invoking
可选的，如图5所示，所述装置还包括：Optionally, as shown in Figure 5, the device further includes:
采集单元34，用于通过多光谱设备采集所述目标生物特征的至少两个图像，其中，各所述图像对应不同的光谱。The
本发明实施例提供的生物特征识别装置中，各个功能模块运行过程中所采用的方法详解可以参见图1-图2方法实施例的对应方法详解，在此不再赘述。In the biometric identification device provided by the embodiment of the present invention, for details of the methods used in the operation of each functional module, refer to the corresponding method details of the method embodiments in FIGS.
进一步的，依据上述实施例，本发明的另一个实施例还提供了一种计算机可读存储介质，所述存储介质包括存储的程序，其中，在所述程序运行时控制所述存储介质所在设备执行图1-图2所述的生物特征识别方法。Further, according to the above embodiment, another embodiment of the present invention further provides a computer-readable storage medium, the storage medium includes a stored program, wherein when the program runs, the device where the storage medium is located is controlled Perform the biometric identification method described in Figures 1-2.
进一步的，依据上述实施例，本发明的另一个实施例还提供了一种电子设备，所述电子设备包括：Further, according to the above embodiment, another embodiment of the present invention also provides an electronic device, the electronic device includes:
存储器，用于存储程序；memory for storing programs;
处理器，耦合至所述存储器，用于运行所述程序以执行图1-图2所述的生物特征识别方法。A processor, coupled to the memory, is configured to run the program to execute the biometric identification method described in FIGS. 1-2 .
在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
可以理解的是，上述方法及装置中的相关特征可以相互参考。另外，上述实施例中的“第一”、“第二”等是用于区分各实施例，而并不代表各实施例的优劣。It can be understood that the relevant features in the above-mentioned methods and apparatuses may refer to each other. In addition, "first", "second", etc. in the above-mentioned embodiments are used to distinguish each embodiment, and do not represent the advantages and disadvantages of each embodiment.
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。Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, device and unit described above may refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.
在此提供的算法和显示不与任何特定计算机、虚拟系统或者其它设备固有相关。各种通用系统也可以与基于在此的示教一起使用。根据上面的描述，构造这类系统所要求的结构是显而易见的。此外，本发明也不针对任何特定编程语言。应当明白，可以利用各种编程语言实现在此描述的本发明的内容，并且上面对特定语言所做的描述是为了披露本发明的最佳实施方式。The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used with teaching based on this. The structure required to construct such a system is apparent from the above description. Furthermore, the present invention is not directed to any particular programming language. It should be understood that various programming languages may be used to implement the inventions described herein, and that the descriptions of specific languages above are intended to disclose the best mode for carrying out the invention.
在此处所提供的说明书中，说明了大量具体细节。然而，能够理解，本发明的实施例可以在没有这些具体细节的情况下实践。在一些实例中，并未详细示出公知的方法、结构和技术，以便不模糊对本说明书的理解。In the description provided herein, numerous specific details are set forth. It will be understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
此外，本领域的技术人员能够理解，尽管在此所述的一些实施例包括其它实施例中所包括的某些特征而不是其它特征，但是不同实施例的特征的组合意味着处于本发明的范围之内并且形成不同的实施例。例如，在权利要求书中，所要求保护的实施例的任意之一都可以以任意的组合方式来使用。Furthermore, those skilled in the art will appreciate that although some of the embodiments described herein include certain features, but not others, included in other embodiments, that combinations of features of different embodiments are intended to be within the scope of the invention within and form different embodiments. For example, in the claims, any of the claimed embodiments may be used in any combination.
本发明的各个部件实施例可以以硬件实现，或者以在一个或者多个处理器上运行的软件模块实现，或者以它们的组合实现。本领域的技术人员应当理解，可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本发明实施例的深度神经网络模型的运行方法、装置及框架中的一些或者全部部件的一些或者全部功能。本发明还可以实现为用于执行这里所描述的方法的一部分或者全部的设备或者装置程序(例如，计算机程序和计算机程序产品)。这样的实现本发明的程序可以存储在计算机可读介质上，或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到，或者在载体信号上提供，或者以任何其他形式提供。Various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the components in the method, apparatus, and framework for running the deep neural network model according to the embodiments of the present invention. some or all functions. The present invention can also be implemented as apparatus or apparatus programs (eg, computer programs and computer program products) for performing part or all of the methods described herein. Such a program implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from Internet sites, or provided on carrier signals, or in any other form.
应该注意的是上述实施例对本发明进行说明而不是对本发明进行限制，并且本领域技术人员在不脱离所附权利要求的范围的情况下可设计出替换实施例。在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的元件或步骤。位于元件之前的单词“一”或“一个”不排除存在多个这样的元件。本发明可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。It should be noted that the above-described embodiments illustrate rather than limit the invention, and that alternative embodiments may be devised by those skilled in the art without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. do not denote any order. These words can be interpreted as names.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210049340.8ACN114550314A (en) | 2022-01-17 | 2022-01-17 | A biometric identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210049340.8ACN114550314A (en) | 2022-01-17 | 2022-01-17 | A biometric identification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114550314Atrue CN114550314A (en) | 2022-05-27 |
Family
ID=81672573
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210049340.8APendingCN114550314A (en) | 2022-01-17 | 2022-01-17 | A biometric identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114550314A (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102385700A (en)* | 2010-09-01 | 2012-03-21 | 汉王科技股份有限公司 | Off-line handwriting recognizing method and device |
| CN103208012A (en)* | 2013-05-08 | 2013-07-17 | 重庆邮电大学 | Light face recognition method |
| CN103605972A (en)* | 2013-12-10 | 2014-02-26 | 康江科技(北京)有限责任公司 | Non-restricted environment face verification method based on block depth neural network |
| CN103714326A (en)* | 2013-12-26 | 2014-04-09 | 江南大学 | One-sample face identification method |
| CN103839275A (en)* | 2014-03-27 | 2014-06-04 | 中国科学院遥感与数字地球研究所 | Method and device for extraction of paths of hyperspectral image |
| CN105160305A (en)* | 2015-08-10 | 2015-12-16 | 中国民航大学 | Finger multi-mode characteristic fusion method |
| CN107644437A (en)* | 2016-07-21 | 2018-01-30 | 宁波舜宇光电信息有限公司 | Color cast detection system and method based on piecemeal |
| CN108460899A (en)* | 2017-02-20 | 2018-08-28 | 深圳怡化电脑股份有限公司 | A kind of characteristics of image recognition methods and device |
| CN111311563A (en)* | 2020-02-10 | 2020-06-19 | 北京工业大学 | Image tampering detection method based on multi-domain feature fusion |
| CN111582228A (en)* | 2020-05-20 | 2020-08-25 | 深圳前海微众银行股份有限公司 | Recognition method, device, device and storage medium of living palm print |
| CN112149687A (en)* | 2020-08-07 | 2020-12-29 | 武汉万集信息技术有限公司 | Method for object recognition |
| CN112633045A (en)* | 2019-10-09 | 2021-04-09 | 华为技术有限公司 | Obstacle detection method, device, equipment and medium |
- 2022
- 2022-01-17CNCN202210049340.8Apatent/CN114550314A/enactivePending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102385700A (en)* | 2010-09-01 | 2012-03-21 | 汉王科技股份有限公司 | Off-line handwriting recognizing method and device |
| CN103208012A (en)* | 2013-05-08 | 2013-07-17 | 重庆邮电大学 | Light face recognition method |
| CN103605972A (en)* | 2013-12-10 | 2014-02-26 | 康江科技(北京)有限责任公司 | Non-restricted environment face verification method based on block depth neural network |
| CN103714326A (en)* | 2013-12-26 | 2014-04-09 | 江南大学 | One-sample face identification method |
| CN103839275A (en)* | 2014-03-27 | 2014-06-04 | 中国科学院遥感与数字地球研究所 | Method and device for extraction of paths of hyperspectral image |
| CN105160305A (en)* | 2015-08-10 | 2015-12-16 | 中国民航大学 | Finger multi-mode characteristic fusion method |
| CN107644437A (en)* | 2016-07-21 | 2018-01-30 | 宁波舜宇光电信息有限公司 | Color cast detection system and method based on piecemeal |
| CN108460899A (en)* | 2017-02-20 | 2018-08-28 | 深圳怡化电脑股份有限公司 | A kind of characteristics of image recognition methods and device |
| CN112633045A (en)* | 2019-10-09 | 2021-04-09 | 华为技术有限公司 | Obstacle detection method, device, equipment and medium |
| CN111311563A (en)* | 2020-02-10 | 2020-06-19 | 北京工业大学 | Image tampering detection method based on multi-domain feature fusion |
| CN111582228A (en)* | 2020-05-20 | 2020-08-25 | 深圳前海微众银行股份有限公司 | Recognition method, device, device and storage medium of living palm print |
| CN112149687A (en)* | 2020-08-07 | 2020-12-29 | 武汉万集信息技术有限公司 | Method for object recognition |
Non-Patent Citations (1)
| Title |
|---|
| 王丽佳、张华、温彬彬: "基于OpenCV的移动机器人目标追踪实现", 31 October 2021, 天津大学出版社, pages: 90 - 93* |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103793690B (en) | A kind of human-body biological biopsy method detected based on subcutaneous haematic flow and application | |
| CN108009472B (en) | Finger back joint print recognition method based on convolutional neural network and Bayes classifier | |
| CN112136140B (en) | Method and apparatus for image recognition | |
| CN111860078B (en) | Method, device, readable storage medium and equipment for silent face liveness detection | |
| CN110443146A (en) | Auth method, device, equipment and readable medium based on bio-identification | |
| CN110738141A (en) | vein identification method, device, equipment and storage medium | |
| CN104102903A (en) | Secondary face recognition method based on SRC | |
| CN112966643A (en) | Face and iris fusion recognition method and device based on self-adaptive weighting | |
| CN108319887A (en) | A kind of identity authentication method and system | |
| CN102722701B (en) | Visual monitoring method and device in fingerprint collection process | |
| WO2025035915A1 (en) | Palm feature processing method and apparatus for identity authentication, and device and medium | |
| KR102205495B1 (en) | Method and apparatus for recognizing finger print | |
| CN107004131A (en) | The method and device of fingerprint recognition | |
| Patil et al. | Fingerprint classification using artificial neural network | |
| CN111222367B (en) | Fingerprint identification method and device, storage medium and terminal | |
| CN103984415B (en) | A kind of information processing method and electronic equipment | |
| CN114612984A (en) | A method and system for face live detection | |
| CN114550314A (en) | A biometric identification method and device | |
| CN106156774B (en) | Image processing method and image processing system | |
| CN113268769A (en) | Method, system and storage medium for processing biological characteristic data based on random confusion | |
| CN102385693B (en) | Finger vein recognition system and method | |
| JP2010277232A (en) | System, method and device for biometrics authentication | |
| WO2018096052A1 (en) | A quick match algorithm for biometric data | |
| Bong et al. | Palm print verification system | |
| CN1257480C (en) | Blood vessel identifying method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information | Address after:Room 263, block B, science and technology innovation center, 128 Shuanglian Road, Haining Economic Development Zone, Haining City, Jiaxing City, Zhejiang Province, 314400 Applicant after:Haining yisiwei IC Design Co.,Ltd. Applicant after:Beijing ESWIN Computing Technology Co.,Ltd. Address before:Room 263, block B, science and technology innovation center, 128 Shuanglian Road, Haining Economic Development Zone, Haining City, Jiaxing City, Zhejiang Province, 314400 Applicant before:Haining yisiwei IC Design Co.,Ltd. Applicant before:Beijing yisiwei Computing Technology Co.,Ltd. | |
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information | Country or region after:China Address after:314400 Building 1, Juanhu Science and Technology Innovation Park, No. 500 Shuiyueting East Road, Xiashi Street, Haining City, Jiaxing City, Zhejiang Province (self declared) Applicant after:Haining Yisiwei Computing Technology Co.,Ltd. Applicant after:Beijing ESWIN Computing Technology Co.,Ltd. Address before:Room 263, block B, science and technology innovation center, 128 Shuanglian Road, Haining Economic Development Zone, Haining City, Jiaxing City, Zhejiang Province, 314400 Applicant before:Haining yisiwei IC Design Co.,Ltd. Country or region before:China Applicant before:Beijing ESWIN Computing Technology Co.,Ltd. |