WO2013075497A1 - 信息采集装置和方法以及身份识别系统和方法 - Google Patents
信息采集装置和方法以及身份识别系统和方法 Download PDFInfo
- Publication number
- WO2013075497A1 WO2013075497A1 PCT/CN2012/077957 CN2012077957W WO2013075497A1 WO 2013075497 A1 WO2013075497 A1 WO 2013075497A1 CN 2012077957 W CN2012077957 W CN 2012077957W WO 2013075497 A1 WO2013075497 A1 WO 2013075497A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- information
- monocular
- bionic unit
- bionic
- unit
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/693—Acquisition
Definitions
- the present invention relates to the field of Internet of Things, and in particular to collecting a plurality of natural biological features for identification, and performing access control and security fields.
- BACKGROUND OF THE INVENTION The starting point of the Internet of Things is goods, people are surrounded by a variety of items, eating and drinking, and work, people will encounter thousands of items. The most intuitive understanding of the Internet of Things is that people need to know which items they want to control, and they want to control everything to play its role.
- the object is the object and the person is the subject.
- the essence of the Internet of Things is mainly embodied in three aspects: First, the characteristics of the Internet, that is, the Internet that needs to be interconnected must be able to achieve interconnection and interconnection; Second, the identification and communication characteristics, that is, the "object” of the Internet of Things must be included It must have the function of automatic identification and object communication (M2M); the third is the intelligent feature, that is, the network system should have the characteristics of automation, self-feedback and intelligent control. Among them, “identification and communication features” and “intelligent features” are inseparable from the identification of objects and related personnel. The objects have the function of automatic identification and object communication, and the Internet of Things platform, objects and personnel can automatically identify each other.
- M2M automatic identification and object communication
- EID technology is an electronic identification technology. From the beginning of computer generation, EID requirements are generated. The use of passwords to verify the identity of computer users is the earliest EID application. Everything that the Internet of Things can identify between users and objects is represented by a specific set of data that represents a digital identity. All authorizations to users and objects are also authorizations for digital identities. How to ensure that the operator operating in digital identity is the legal owner of the digital identity, that is, to ensure that the physical identity of the operator corresponds to the digital identity, the EID service is to solve this problem, as the first pass of the Internet of Things, EID services play a pivotal role. The means of identity authentication in the Internet of Things is consistent with the real world.
- the basic methods of user identity authentication can be divided into three types: telling what you know; showing what you have; providing unique biometrics.
- the human natural recognition process shown in Figure 1 is a process in which humans do not recognize each other by means of tools.
- the other party is used to obtain facial expressions, expressions, costumes (types, patterns, colors, sizes, etc.), hairstyles, postures, etc., and then The characteristics of the memory are compared. Finally, the identity of the other party is judged, and other information of the other party is recalled according to the identity information.
- two levels of identity authentication mechanisms are described: simple authentication, strong authentication. Simple authentication is the identification of IDs and passwords.
- the user enters the ID and password in the client, and the server is responsible for identifying the identity.
- the transmission of the password from the client to the server is usually encrypted with a symmetric key.
- Strong authentication based on PKI (Public Secret) Key infrastructure) is identified by digital certificates and digital signatures.
- the digital certificate and private key are usually stored in a USB Key or Smart Card.
- the user inserts a USB Key or Smart Card on the client side, and the server side is responsible for identifying the identity.
- the transmission of the password from the client to the server is encrypted with an asymmetric key. Strong authentication mechanism can also use passwords, because the digital certificate file, USB Key, and Smart Card are saved by the user and are easily stolen.
- the narrow password refers to static passwords and dynamic passwords.
- the broad passwords include digital certificates and biometrics. IDs and passwords can be stored in smart cards or USB keys.
- biometrics including speech, face, skin, iris, retina, body, keyboard, signature, etc.
- natural and undetected biometrics have attracted attention, such as: face, voice, skin, iris and body.
- the so-called naturalness means that the recognition method is the same as that used by humans (or even other organisms) for individual identification.
- Undetected features are also important for an identification method, which makes the recognition method unobjectionable and not easily fooled because it is not easily noticeable.
- the following are examples of five biological features that are both natural and undetectable.
- Face recognition is the most natural and undetected feature. Face features contain expressions and shapes that affect each other. Although face recognition has many advantages that other recognitions can't match, it has many flaws. Face recognition is considered to be one of the most difficult research topics in the field of biometrics and even in the field of artificial intelligence.
- Speech recognition speech recognition The recording device continuously measures and records the waveform and changes of the sound, and matches the collected sound on the scene with the registered sound template to determine the user's identity. This recognition technology is recognized for technical problems. not tall. The disadvantages of speech recognition are as follows:
- the skin texture is the texture formed by the skin of the primate creatures. It can also refer to the impressions printed on the objects.
- the detailed feature points of the texture have a starting point, an ending point, a joint point, and a bifurcation point. Since each person's fingerprint is not the same, the skin texture of different parts of the same person is different.
- the skin pattern recognition is identified by comparing the differences of these details. Skin pattern recognition has a long history as a recognition technology, and it is possible to reliably confirm a person's identity.
- the texture recognition technology represented by fingerprint recognition has the following disadvantages: (1) Currently widely used identification devices are not accurate and have a resolution of 500 DPI.
- Capacitive fingerprint identification instrument Some identification devices are contactive, and some people think that they are unhygienic and refuse to use, for example: Capacitive fingerprint identification instrument.
- Skin marks are often not recognized when sweating or being contaminated, and need to be cleaned or repeatedly identified to pass.
- Iris recognition The formation of the iris is determined by genetics, and the expression of the human body determines the shape, physiology, color and overall appearance of the iris. Iris recognition technology can achieve very good accuracy. Even if all human iris information is entered into a single data, the possibility of falsehood and rejection is quite small, and the accuracy is higher than any other biometric authentication technology. Dozens of orders of magnitude. The disadvantages of iris recognition are:
- the camera scans the user's eyes at a close distance, which is an intrusive recognition method, and some users are more disgusted.
- Body posture is the posture and shape of the body.
- the body posture contains a lot of contents, such as: gait, standing posture, sitting posture, gestures and costume dressing.
- gait refers to a complex behavioral characteristic when people walk.
- the main feature extracted by gait recognition is the movement of each joint of the human body. Although gait is not different for everyone, it also provides sufficient information to identify a person.
- the input of gait recognition is a sequence of walking video images. However, due to the large amount of data in the sequence image, the computational complexity of gait recognition is relatively high and it is difficult to handle.
- insects are popular subjects in the scientific and engineering circles. They can give you an inspiration for information collection methods.
- the collection objects are natural and imperceptible.
- Insect eyes include monocular and compound eyes, and monocular and monocular and lateral monocular. Insects respond to changes in external light through monocular and compound eyes, and perform activities such as foraging, courtship, orientation, dormancy, and diapause.
- the compound eye is the main visual organ of the insect and usually occupies a prominent position on the head of the insect.
- the compound eyes of most insects are round, oval or kidney-shaped.
- the compound eye is composed of many hexagonal small eyes, each of which has the same basic structure as a single eye.
- the monocular eye can increase the response of the compound eye to the light stimuli. Some insects can also distinguish the contour of the object, the color of the light and the movement of the near object. Some studies have concluded that the monocular eye is several times more sensitive to light than the compound eye, so information from the monocular can be used to adjust the response of the compound eye to stimuli. Studies have also shown that bees searching for food under low light intensity usually have large monocular ones, and monocular ones are also related to flying.
- the functions of bees, ants and scorpion monoculars also involve the orientation of flight; To the control of insect white mites, the detection of polarized light, the secretion of the neurosecretory system, etc., it can be said that the monocular is a physiology rhythm receptor. Biologists have also found that the monocular eye can sense color, shape, distance, motion, and polarized light, so that insects with side monoculars acquire the corresponding capabilities.
- the side monocular can also be imaged, but because the number of photoreceptors under the lens is too small, the spatial extent of the receiving area between each photoreceptor does not overlap 50%, and the field of view cannot completely cover the surrounding environment, so only a rough mosaic can be formed. image.
- Embodiments of the present invention provide an information collecting apparatus and method, and an identity recognition system and method, which utilize the bionics of an insect visual organ to realize intelligent collection and recognition of target biological features.
- An embodiment of the present invention provides an information collecting apparatus, including: a back monocular bionic unit, a side monocular bionic unit, and a compound eye bionic unit; the back monocular bionic unit is configured to collect environmental parameter information; and the side monocular bionic unit is used for collecting Target contour information; the compound eye bionic unit is configured to collect dynamic image information according to the environmental parameter information and the target contour information.
- the embodiment of the invention further provides an identity recognition system, comprising the above-mentioned information collection device for collecting physical identity information, and an information recognition device for identifying the physical identity information.
- An information collection method includes the following processes: collecting environmental parameter information by a single-eye bionic unit; collecting target contour information by a side single-eye bionic unit; and passing the environmental parameter information and target contour information, and The compound eye bionic unit collects dynamic image information.
- An identification method according to an embodiment of the present invention, the processing step of collecting the physical identity information by the information collection method, and the processing step of performing identification according to the physical identity information.
- the information collection and recognition technology proposed by the embodiment of the invention has strong universality, and is particularly suitable for the Internet of Things. Due to the physical and virtual "things" and “people” throughout the Internet of Things, there are interfaces for identity, physical attributes, virtual features, and intelligence.
- FIG. 1 is a flow chart of a human natural recognition process
- FIG. 2 is a flow chart of an insect vision process
- FIG. 3 is a schematic structural view of an identification system according to a preferred embodiment of the present invention
- FIG. 5 is a flow chart of a method for collecting information in a preferred embodiment of the present invention
- FIG. 6 is a flowchart of an identity recognition process in a preferred embodiment of the present invention.
- the present embodiment draws on the research results of insect visual organs and performs simplified processing of bionics to complete the mutual cooperation of multiple organs.
- the coordinated working principle of the monocular and lateral monocular and compound eyes in the insect visual organ is respectively used, as shown in Fig. 2.
- the use of this multi-visual organ interaction mechanism can solve the shortcomings of traditional identification methods.
- the traditional identification method is not suitable for the Internet of Things that contains the masses and their vast items, because traditional identification methods are either not natural and not Perceptuality, either the rejection rate and the refusal rate are relatively high.
- the information collecting apparatus of the embodiment of the invention comprises a monocular bionic unit, a side monocular bionic unit and a compound eye bionic unit, which respectively play the monocular, lateral monocular and compound eyes of the insect.
- the single-eye bionic unit consists of several environmental parameter sensors, such as: light intensity sensor, humidity sensor, temperature sensor, odor sensor, and the single-eye bionic unit can achieve the purpose by using one. However, if the general direction judgment ability is increased, it needs to be increased. Quantity.
- the side monocular bionic unit consists of at least one camera, and the resolution of the camera is not required to be too high, such as a 100 DPI CMOS sensor.
- the compound eye bionic unit uses a camera array, which is placed on a curved surface by a large number of cameras.
- the camera preferably uses a higher resolution sensor, such as a 600DPI CCD sensor.
- the surface can be placed in a plane or a spherical surface. The selection is based on sharpness and motion.
- the rate, camera area and camera angle, the compound eye bionic unit can also be replaced by a commercial compound eye bionic camera.
- a control unit may be disposed in the information collection device, and the back monocular bionic unit, the side monocular bionic unit, and the compound eye bionic unit are respectively connected, and the signals collected by the bionic unit sensors are received, and may be further saved.
- the identity recognition system of the embodiment of the present invention includes an information recognition device in addition to the information collection device.
- the information collection device is configured to collect physical identity information
- the information identification device is configured to identify and process the collected physical identity information.
- the information identifying apparatus of the embodiment of the present invention includes: an identification server and an information database.
- the information database stores digital identity information
- the identification server is configured to perform the identification process of the collected physical identity information in the information database.
- the information database can be divided into: identity database, biometric database, identity database to store the required identity information, biometric database to store the required biometric information, where the identity and biometric information may not be someone's information, but some All information of a group, when the business requirement only needs to identify whether the target is a certain person, the database only needs to save the person's information, and when the business requires to find the identification target in a certain group, the database needs to save all the information of the group. For example: When someone enters a subway station after buying a ticket, the database needs to save information about all the people who bought the ticket and did not have a stop.
- the identification server is responsible for comparing the information collected by the information collection device with the information stored in the database.
- the identification server needs to consider the matching of large-scale biometric information to improve the response speed, for example: using cloud computing instead of a single server software and hardware.
- the access control device can also be set to implement access control according to the recognition result.
- the embodiment of the present invention is applied to the field of the Internet of Things, and an important role is to solve the problem of identity recognition in a smart city, which is to achieve intelligentization and automation of the city, and to help determine the legal identity of users and their items.
- the information collecting device 31 and the information identifying device 32 are connected via a communication network, and the citizens are direct users of the system (direct users).
- the information identifying apparatus includes: an identification server cluster 321, an identity database 322, and a biometric database 323.
- the specific structural design of the information collecting device 31 is shown in FIG. 4 .
- the information collecting device is composed of a back monocular bionic unit 41, a side monocular bionic unit 42, a compound eye bionic unit 43, and a control unit 44.
- the single-eye bionic unit 41 has one, located directly in front, and is composed of a light intensity sensor, a humidity sensor, a temperature sensor, and an odor sensor, and the output signal of the sensor is a digital signal.
- the compound eye bionic unit 43 adopts an image capturing array, and is arranged in a fan shape of 270 degrees by 11 CCD sensors with a resolution of 1200 DPI.
- the compound eye bionic unit 43 also includes a camera array controller to control the camera array.
- the control unit 44 is connected to the back monocular bionic unit 41, the side monocular bionic unit 42, and the compound eye bionic unit 43, respectively, and receives digital signals of the sensor (for receiving information of each unit) and stores them.
- the embodiment of the present invention provides an identity recognition system, including the foregoing information collection device for collecting physical identity information, and an information recognition device for identifying physical identity information.
- the information identifying apparatus comprises: an identification server, an information database; the information database is configured to store digital identity information; and the identification server is configured to perform the physical identity information in digital identity information stored in the information database Identification. As shown in FIG.
- the information collection process of the information collection device includes the following steps: First, collecting environmental parameters, such as light intensity, humidity, temperature, odor, etc., second step, positioning the target, and collecting the target contour, third Step, collect dynamic images.
- the communication network of this embodiment is a heterogeneous communication network constructed by a smart city, and is connected by a high-speed IP backbone network, a mobile communication network, a WIFI wireless network, a sensor network and the like connected by a gateway.
- the information identification device 32 of the present embodiment is composed of an identification server cluster 321, an identity database 322, and a biometric database 3223.
- the identification server cluster 321 is composed of three or more servers, and is responsible for processing environmental parameters collected by the information collection device.
- Detecting data and moving images extracting feature values, where the eigenvalues include six kinds of natural biometric values such as forehead lining, iris, face shape, facial expression, gait, and costume dressing, as well as gender and height.
- the eigenvalues in addition, the identification server cluster is also responsible for the verification of identity information and the verification of biometric values.
- the verification steps are: (1) Extract data from the identity database, compare it with the gender and height of the detected target, and filter out the match.
- condition record the “condition” here includes two cases: completely equal, the error is within a certain range, an example of the latter is that the detected target height is 170 cm, and the setting error is 1 cm, which is consistent with "
- condition "record” is a record of all heights ranging from 169 to 171 cm.
- the information acquisition device acquires environmental parameters (light intensity, humidity, temperature, odor, etc.);
- the information collecting device automatically sets initial detection conditions (area, angle range, speed range, sound intensity range) according to environmental parameters;
- the information collecting device continuously obtains environmental parameters, and judges whether there is human existence according to the change thereof. If any human exists in the vicinity, the next step is entered, otherwise, standby;
- the information collecting device detects the target position, the contour, the sound intensity and the motion track, and further determines whether the detection condition is met, and if yes, proceeds to the next step, otherwise, stands by;
- the information collecting device captures a dynamic image centered on the detected target
- the information collecting device stores the moving image data, the environmental parameters, the detection data, and determines whether to upload the data to the information identifying device according to the service;
- the information recognition device processes the environment parameters, and the processing steps are: 1 acquiring the service type, 2 acquiring the environment parameter transformation algorithm according to the service type, and 3 calculating the new environment parameter according to the algorithm;
- the information recognition device processes the detection data, and the processing steps are: 1 acquiring the service type, 2 acquiring the detection data transformation algorithm according to the service type, and 3 calculating the new detection data according to the algorithm;
- the information recognition device processes the moving image data, and the processing steps are: 1 acquiring the service type, 2 acquiring the dynamic image data feature extraction algorithm according to the service type, 3 correcting the dynamic image according to the new environmental parameter, 4 correcting the dynamic according to the new detection data Image, 5 calculates biometric values according to an algorithm;
- the information recognition device compares the registered biometric template with the biometric value just obtained, and if it is met, proceeds to the next step; otherwise, it warns or prohibits the action of the detected object (depending on the business);
- the information recognition device acquires an access control rule, for example: the authority or security level of the detected target, and hands it over to other systems for processing.
- an access control rule for example: the authority or security level of the detected target, and hands it over to other systems for processing.
Abstract
本发明公开了一种信息识别装置和方法以及一种身份识别装置和方法，通过设置背单眼仿生单元、侧单眼仿生单元和复眼仿生单元，来完成多仿生单元的相互协同工作，实现目标生物特征的智能化采集和识别。采用本发明可于信息网络无缝整合，对于"物"和"人"具有的覆盖面广、复杂多变等特性，实现物理身份信息的采集和身份识别。
Description
信息采集装置和方法以及身份识别系统和方法 技术领域 本发明涉及物联网领域， 具体地说， 涉及采集多种自然生物特征进行身份识别， 并进行访问控制和安全领域。 背景技术 物联网的起点是物品， 人周围是各种各样的物品， 吃穿住行， 还有工作， 人们会 遇到成千上万种物品。 对物联网最直观的理解就是， 人们需要知道哪些物品， 希望控 制所有物品发挥其作用， 物品是客体， 人是主体。 物联网的本质概括起来主要体现在 三个方面： 一是互联网特征， 即对需要联网的物一定要能够实现互联互通的互联网络； 二是识别与通信特征， 即纳入物联网的"物"一定要具备自动识别与物物通信 （M2M) 的功能； 三是智能化特征， 即网络系统应具有自动化、 自我反馈与智能控制的特点。 其中， "识别与通信特征"和"智能化特征"都离不开物体和相关人员的身份识别， 物体 具备自动识别与物物通信的功能， 物联网平台、 物体、 人员相互能够自动识别身份。
EID技术就是电子身份识别技术， 从计算机产生之初， 就产生了 EID需求， 使用 口令来验证计算机使用者的身份就是最早的 EID应用。 物联网能识别用户和物体的一 切信息都是用一组特定的数据来表示的， 这组特定的数据代表了数字身份， 所有对用 户和物体的授权也是针对数字身份的授权。 如何保证以数字身份进行操作的操作者就 是这个数字身份合法拥有者，也就是说保证操作者的物理身份与数字身份相对应， EID 服务就是为了解决这个问题， 作为物联网的第一道关口， EID服务有着举足轻重的作 用。 在物联网中身份认证手段与真实世界中一致。 在真实世界， 对用户身份认证的基 本方法可以分为三种： 说出所知道的信息； 展示所拥有的东西； 提供独一无二的生物 特征。 如图 1所示的人类自然识别过程是人类不借助工具相互识别的过程， 通常先打 量对方， 获取脸型、 表情、 服饰 （类型、 花纹、 颜色、 大小等）、 发型、 体态等信息， 然后从自己记忆的特征对比， 最后， 判断出对方身份， 根据身份信息回忆对方的其他 信息。 在 X.509标准中， 描述了两个级别的身份认证机制： 简单认证、 强认证。 简单认 证就是 ID和口令的识别。 用户在客户端中输入 ID和口令， 服务器端负责识别身份。 口令从客户端传送服务器端的传输通常采用对称密钥加密。 强认证基于 PKI (公共密
钥基础设施）采用数字证书、数字签名进行识别。数字证书和私钥通常存储在 USB Key 或 Smart Card中，用户在客户端插入 USB Key或 Smart Card,服务器端负责识别身份。 口令从客户端传送服务器端的传输采用非对称密钥加密。强认证机制也可以使用口令， 因为数字证书文件、 USB Key、 Smart Card的保存由用户负责， 容易被窃用， 所以， 采用强双因子认证方法是非常可行的识别认证手段， 即同时使用口令和证书。 狭义的口令指静态密码和动态口令， 广义的口令就包括了数字证书和生物特征， ID和口令可存储在智能卡或者 USB Key中。 但是， 目前生物特征的研究领域非常多， 主要包括语音、 人脸、 皮纹、 虹膜、 视网膜、 体态、 敲击键盘、 签字等。 针对物联网， 具有自然性和不被察觉性的生物特征引起了重视， 例如： 人脸、 语音、 皮纹、 虹膜和 体态。 所谓自然性， 是指该识别方式同人类 （甚至其他生物） 进行个体识别时所利用 的生物特征相同。 不被察觉的特点对于一种识别方法也很重要， 这会使该识别方法不 令人反感， 并且因为不容易引起人的注意而不容易被欺骗。 下面举例说明五种具有自 然性和不被察觉性的生物特征。
1. 人脸识别 人脸识别最能体现自然性和不被察觉的特点。 人脸特征包含表情和形状， 两者互 相影响。虽然人脸识别有很多其他识别无法比拟的优点，但是它本身也存在许多缺陷。 人脸识别被认为是生物特征识别领域， 甚至人工智能领域最困难的研究课题之一。
2. 语音识别 语音识别利用录音设备不断地测量、 纪录声音的波形和变化， 将现场采集到的声 音同登记过的声音模板进行匹配， 从而确定用户的身份， 这种识别技术因为技术问题 识别精度不高。 语音识别的缺点如下：
( 1 ) 声音会随着音量、 速度和音质的变化而影响到采集与比对的结果。
(2) 很容易录制好的声音来欺骗。
3. 皮纹识别 皮纹是灵长类生物凹凸的皮肤所形成的纹路， 也可指这些纹路在物体上印下的印 痕。 纹路的细节特征点有起点、 终点、 结合点和分叉点。 由于每个人的指纹并不相同， 同一人的不同部位的皮纹也不一样， 皮纹识别就是通过比较这些细节特征的区别来进 行鉴别。皮纹识别作为识别技术已经有很长的历史了，可以可靠地确认一个人的身份。 但是， 在现实应用中， 以指纹识别为代表的皮纹识别技术有如下缺点：
( 1 ) 现在普遍采用的识别设备精度不高， 分辨率在 500 DPI。
(2)纹理痕迹存在被复制的可能性， 例如： 指纹容易被他人获取且利用一定科技 含量的手段可以克隆指纹。
(3 )某些人某些部位的皮纹很难成像， 增大了拒真率， 例如： 矿工的指纹特征就 很少。
(4)有些识别设备是接触性的， 有人认为不卫生拒绝使用， 例如： 电容式指纹识 别仪。
( 5 ) 出汗或被污染时常常无法识别皮纹， 需要洁净或反复识别才能通过。
4. 虹膜识别 虹膜的形成由遗传基因决定， 人体基因表达决定了虹膜的形态、 生理、 颜色和总 的外观。 虹膜识别技术可以达到十分优异的准确度， 即使全人类的虹膜信息都录入到 一个数据中， 出现认假和拒假的可能性也相当小， 比其他任何生物认证技术的精确度 高几个到几十个数量级。 虹膜识别的缺点是：
( 1 ) 虹膜识别对使用者有较高要求， 眼睛必须对准摄像头
(2)摄像头近距离扫描用户的眼睛，是一种侵入式识别方式，有的用户比较反感。
5. 体态识别 体态是身体的姿态和形态， 体态包含非常多的内容， 例如： 步态、 站姿、 坐姿、 手势和服饰装扮。 作为典型的一种体态， 步态是指人们行走时的一种复杂行为特征。 步态识别主要提取的特征是人体每个关节的运动。 尽管步态不是每个人都不相同的， 但是它也提供了充足的信息来识别人的身份。 步态识别的输入是一段行走的视频图像 序列。 但是， 由于序列图像的数据量较大， 因此步态识别的计算复杂性比较高， 处理 起来也比较困难。 上述这些生物特征都具有自然性和不可觉察性， 故生物自然而然地使用视觉器官 捕捉这些生物特征。 依据仿生学， 昆虫的视觉器官是科学界和工程界人士的研究热门 对象， 它们能给大家一种信息采集方法启示， 采集对象是具有自然性和不可觉察性的 生物特征。
昆虫的眼睛包括单眼与复眼， 单眼又有背单眼与侧单眼之分。 昆虫通过单眼与复 眼对外界光的变化做出反应， 进行觅食、 求偶、 定向、 休眠、 滞育等活动。 复眼是昆虫的主要视觉器官， 通常在昆虫的头部占有突出的位置。 多数昆虫的复 眼呈圆形、 卵圆形或肾形。 复眼是由许多六角形的小眼组成的， 每个小眼与单眼的基 本构造相同。 复眼的体积越大， 小眼的数量就越多， 看东西的视力也就越强。 背单眼能增加复眼感受光线剌激的反应， 有些昆虫的侧单眼还能辨别物体轮廓、 光的颜色和近处物体的移动。 有人研究得出结论： 背单眼对光的敏感性比复眼要高好 几倍， 因此来自于背单眼的信息可用于调整复眼对于剌激的反应。 还有研究表明， 在 低光照强度下搜索食物的蜜蜂通常具有大的背单眼， 背单眼还与飞翔有关， 蜜蜂、 蚂 蚁和蜻蜓背单眼的功能还涉及到飞行的定向； 背单眼的功能还涉及到对昆虫白昼活动 的控制、偏振光的探测、神经分泌系统的分泌等， 可以说背单眼是生理节奏的感受器。 生物学家还发现， 侧单眼能够感知颜色、 形状、 距离、 运动和偏振光等， 因此具 有侧单眼的昆虫就获得了相应的能力。 侧单眼也能够成像， 但因透镜下光感受器数目 太少， 每个光感受器间接受区的空间范围有 50 %不重叠， 其视野不能完全覆盖周围的 环境， 因此只能形成一个粗糙的镶嵌性图像。具侧单眼的幼虫能感知周围环境的反差， 这可以通过它们具有的一些趋性反应得到验证。 根据以上的昆虫视觉器官研究成果， 昆虫视觉过程是昆虫众多视觉器官相互配合 的过程， 昆虫视觉器官包括复眼、 背单眼、 侧单眼等。 相关技术中还没有公开利用昆虫视觉器官的研究成果来进行生物特征信息采集， 并将其用于身份识别的技术。 发明内容 本发明实施例提供了一种信息采集装置和方法， 以及身份识别系统和方法， 利用 昆虫视觉器官的仿生， 实现目标生物特征的智能化采集和识别。 本发明实施例提出了一种信息采集装置， 包括： 背单眼仿生单元、 侧单眼仿生单 元和复眼仿生单元； 所述背单眼仿生单元用于采集环境参数信息； 所述侧单眼仿生单 元用于采集目标轮廓信息； 所述复眼仿生单元用于根据所述环境参数信息和目标轮廓 信息采集动态图像信息。 本发明实施例还提出了一种身份识别系统， 包括上述的用于采集物理身份信息的 信息采集装置， 以及用于对所述物理身份信息进行识别的信息识别装置。
本发明实施例提出的一种信息采集方法， 包括以下处理过程： 通过背单眼仿生单 元采集环境参数信息； 通过侧单眼仿生单元采集目标轮廓信息； 根据所述环境参数信 息和目标轮廓信息， 并通过复眼仿生单元采集动态图像信息。、 本发明实施例提出的一种身份识别方法， 上述的信息采集方法采集物理身份信息 的处理步骤， 还包括根据所述物理身份信息进行识别的处理步骤。 本发明实施例提出的信息采集和识别技术具有较强的普适性，尤其适合于物联网。 由于遍布物联网的物理的和虚拟的 "物"和"人"， 具有身份标识、 物理属性、 虚拟的特 性和智能的接口。 采用本发明可于信息网络无缝整合， 对于"物"和"人"具有的覆盖面 广、 复杂多变等特性， 实现物理身份信息的采集和身份识别。 附图说明 图 1为相关技术是人类自然识别处理流程图； 图 2为昆虫视觉处理流程图； 图 3为本发明优选实施例的身份识别系统结构示意图； 图 4为本发明优选实施例中的信息采集装置结构示意图； 图 5为本发明优选实施例中信息采集方法流程图； 以及 图 6为本发明优选实施例中的身份识别处理流程图。 具体实施方式 本实施例借鉴昆虫视觉器官的研究成果， 并进行仿生学的简化处理， 来完成多器 官的相互协同工作。分别利用昆虫视觉器官中背单眼和侧单眼和复眼的协调工作原理， 如图 2所示。 首先， 使用背单眼感应光的强度， 分辨白天和黑夜， 确定方向， 然后， 使用侧单眼辨别物体轮廓， 感应运动轨迹， 最后根据背单眼和侧单眼获取的状态来调 整复眼（方向、 焦距、 光圈）， 获取清晰的动态视觉图像。 利用这种多视觉器官相互协 同的机制可以解决传统身份识别方法的缺陷， 传统身份识别方法不适合包含大众及其 浩瀚如海的物品的物联网， 因为传统身份识别方法要么不具备自然性和不可察觉性， 要么拒真率和拒假率都比较高。
本发明实施例的信息采集装置包括背单眼仿生单元、 侧单眼仿生单元和复眼仿生 单元， 分别扮演昆虫的背单眼、 侧单眼和复眼。 背单眼仿生单元由若干个环境参数传 感器组成， 例如： 光强度传感器、 湿度传感器、 温度传感器、 气味传感器， 背单眼仿 生单元使用一个就能达到目的， 但是， 如果增加大致方向判断能力， 就需要增加数量。 侧单眼仿生单元由至少一个摄像头组成， 摄像头的分辨率不要求太高， 例如 100DPI 的 CMOS传感器。 当侧单眼仿生单元数量大于等于二时， 可以获取目标位置和大致运 动轨迹。 复眼仿生单元采用摄像阵列， 由大量摄像头按曲面摆放， 摄像头最好采用分 辨率较高的传感器，例如 600DPI的 CCD传感器， 曲面摆放的方式可以是平面和球面， 选择依据是清晰度、 运动速率、 摄像区域和摄像角度， 复眼仿生单元也可以采用商业 的复眼仿生摄像机代替。 此外， 还可以在信息采集装置中设置控制单元， 分别连接背 单眼仿生单元、侧单眼仿生单元和复眼仿生单元，接收各仿生单元传感器采集的信号， 还可以进一步将其保存起来。 本发明实施例的身份识别系统除了包括信息采集装置之外，还包括信息识别装置。 其中信息采集装置用来采集物理身份信息， 信息识别装置用来对采集的物理身份信息 进行识别处理。 本发明实施例的信息识别装置包括： 识别服务器和信息数据库。 信息数据库存储 数字身份信息，识别服务器用于将采集的物理身份信息在信息数据库中进行识别处理。 信息数据库可以分为： 身份数据库、 生物特征数据库， 身份数据库存储所需的身份信 息， 生物特征数据库存储所需的生物特征信息， 这里的身份和生物特征信息可能不是 某个人的信息， 而是某个群体的所有信息， 当业务要求仅仅需要识别目标是否某个人 时，数据库仅需要保存这个人的信息， 而当业务要求在某个群体里面寻找识别目标时， 数据库需要保存这个群体的所有信息， 例如： 某个人买票后进入地铁车站时， 数据库 需要保存所有买票且没有进站的所有人的信息。 识别服务器负责对比信息采集装置采 集的信息和数据库存储的信息， 当数据库庞大时， 识别服务器需要考虑大规模生物特 征信息的匹配， 以提高响应速度， 例如： 采用云计算来代替单台服务器软件和硬件。 此外， 根据身份识别的应用情况， 还可以设置访问控制设备， 实现根据识别结果 的访问控制。 下面结合附图， 并通过具体实施例对本发明实施例的实现进行详细说明。 本实施 例应用于智慧城市。 在政府的推动下， 物联网广泛应用于智慧城市的各个领域， 从而 解决目前城市遇到的诸多问题， 打破政府部门之间的围墙， 提高政府办事效率， 为市 民办实事， 提供群众的满意度。 当前城市遇到了诸如老龄化、 城市化、 贫富差距大等
众多问题。 本发明实施例运用于物联网领域， 重要作用就是解决智慧城市的身份识别 问题， 是城市达到智能化和自动化， 帮助确定用户及其物品的合法身份。 如图 3所示的系统结构，信息采集装置 31和信息识别装置 32通过通讯网络连接， 市民都是系统直接使用者 （直接用户）。 信息识别装置中包括： 识别服务器集群 321、 身份数据库 322和生物特征数据库 323。 其中信息采集装置 31的具体结构设计如图 4所示。信息采集装置由背单眼仿生单 元 41、侧单眼仿生单元 42、 复眼仿生单元 43和控制单元 44组成。本实施例中背单眼 仿生单元 41有一个， 位于正前方， 由一个光强度传感器、 一个湿度传感器、一个温度 传感器、 一个气味传感器组成， 传感器的输出信号为数字信号。 侧单眼仿生单元 42 有两个，位于左右两侧，分别由两个 CCD传感器组成， CCD传感器的分辨率为 300DPI。 复眼仿生单元 43采用摄像阵列， 由 11个 CCD传感器按 270度扇形摆放， 分辨率为 1200DPI。 复眼仿生单元 43还包括摄像阵列控制器， 以便控制摄像阵列。 控制单元 44 分别连接背单眼仿生单元 41、 侧单眼仿生单元 42和复眼仿生单元 43， 接收传感器的 数码信号 （用于接收各单元的信息）， 并保存起来。 本实施例实施例提出了一种身份识别系统， 包括上述的用于采集物理身份信息的 信息采集装置， 以及用于对物理身份信息进行识别的信息识别装置。 优选地， 该信息识别装置包括： 识别服务器、 信息数据库； 所述信息数据库用于 存储数字身份信息； 所述识别服务器用于将所述物理身份信息在所述信息数据库存储 的数字身份信息中进行身份识别。 如图 5所示， 信息采集装置的信息采集流程包括以下步骤： 第一步， 采集环境参 数， 例如光强， 湿度， 温度， 气味等， 第二步， 定位目标， 并采集目标轮廓， 第三步， 采集动态图像。 本实施例的通讯网络是智慧城市构建的异构通讯网络，通过网关连接的高速 IP骨 干网、 移动通讯网、 WIFI无线网络、 传感网等相互连接而成。 本实施例的信息识别装置 32由识别服务器集群 321、 身份数据库 322、 生物识别 数据库 3223组成， 其中， 识别服务器集群 321由三台及三台以上的服务器组成， 负责 处理信息采集装置采集的环境参数、 检测数据和动态图像， 提取特征值， 此处的特征 值分别包含额头皮纹、 虹膜、 脸部形状、 脸部表情、 步态和服饰装扮六种自然生物特 征值以及性别、 身高两种普通特征值， 除此之外， 识别服务器集群还负责身份信息的 核对和生物特征值核对工作， 核对步骤为：
( 1 )从身份数据库提取数据， 和被检测目标的性别、 身高逐一对比， 筛选出符合
"条件 "的记录， 此处的 "条件 "包括两种情况： 完全相等、 误差在一定范围内， 后者的 一个例子是， 被检测目标身高为 170厘米， 设定误差为 1厘米， 符合"条件"的记录则 是所有身高在 169到 171厘米范围内的记录。
(2)从生物特征数据库提取数据， 和被检测目标的生物特征值逐一对比， 筛选出 符合"条件"的记录， 此处的 "条件 "指符合度在一定范围内， 假设符合度为 99.9%， 则 符合判断的公式为：
(被检测目标和记录的生物特征点相同的数量） ÷记录的生物特征点 > 99.9%
(3 ) 如果符合条件的记录数量为 1， 则认为这条记录就是被检测人的身份信息； 如果符合条件的记录数量大于 1， 则设置更严格的"条件"缩小记录数量， 否则， 设置 更宽松的"条件"增加记录数量， 经过这样的自动处理， 如果符合条件的最终记录数量 为 1， 则认为这条记录就是被检测人的身份信息， 如果符合条件的最终记录数量为 0， 则认为被检测人不在数据库内， 因此认为被检测人不合法。 如图 6所示， 本实施例的具体身份识别处理过程， 描述如下：
( 1 ) 信息采集装置获取环境参数 （光强度、 湿度、 温度、 气味等等）；
(2)信息采集装置根据环境参数自动设定初始检测条件（区域、 角度范围、 速度 范围、 声音强度范围）；
(3 )信息采集装置持续获取环境参数， 并根据其变化判断是否有人类存在， 若有 人类在附近存在， 则进入下一步， 否则， 待机；
(4)信息采集装置检测目标位置、 轮廓、 声音强度和运动轨迹， 并进一步判断是 否符合检测条件， 若符合则进入下一步， 否则， 待机；
( 5 ) 信息采集装置捕获以被检测目标为中心的动态图像；
(6)信息采集装置存储该动态图像数据、 环境参数、 检测数据， 并根据业务决定 是否上传这些数据到信息识别装置；
(7)信息识别装置处理环境参数， 处理步骤为： ①获取业务类型， ②按业务类型 获取环境参数变换算法， ③根据算法计算新的环境参数；
( 8 )信息识别装置处理检测数据， 处理步骤为： ①获取业务类型， ②按业务类型 获取检测数据变换算法， ③根据算法计算新的检测数据；
( 9)信息识别装置处理动态图像数据， 处理步骤为： ①获取业务类型， ②按业务 类型获取动态图像数据特征提取算法， ③根据新的环境参数修正动态图像， ④根据新 的检测数据修正动态图像， ⑤根据算法计算生物特征值；
( 10) 信息识别装置使用刚获得的生物特征值来对比已登记的生物特征模板， 如 果符合，则进入下一步，否则，对被检测目标提出警告或者禁止行动（根据业务而定）;
( 11 )信息识别装置获取访问控制规则， 例如： 被检测目标的权限或者安全等级， 并交给其他系统处理。 是结合具体的实施方式对本发明所作的进一步详细说明， 不能认定本发明的具体 实施只局限于这些说明； 因此， 对于本发明所属技术领域的普通技术人员来说， 在不 脱离本发明构思的前提下， 还可以做出若干简单推演或替换， 都应当视为属于本发明 的保护范围。
Claims
1. 一种信息采集装置， 包括： 背单眼仿生单元、侧单眼仿生单元和复眼仿生单元;
所述背单眼仿生单元用于采集环境参数信息；
所述侧单眼仿生单元用于采集目标轮廓信息；
所述复眼仿生单元用于根据所述环境参数信息和目标轮廓信息采集动态图 像信息。
2. 根据权利要求 1所述的信息采集装置， 其中， 所述背单眼仿生单元包括下列环 境参数传感器中的至少一种： 光强度传感器、 湿度传感器、 温度传感器、 气味 传感器。
3. 根据权利要求 1所述的信息采集装置， 其中， 所述侧单眼仿生单元包括第一图 像传感器， 所述第一图像传感器用于采集目标轮廓信息。
4. 根据权利要求 3所述的信息采集装置， 其中， 所述侧单眼仿生单元包括至少两 个第一图像传感器， 所述各第一图像传感器配合用于采集目标位置信息和运动 轨迹信息。
5. 根据权利要求 1所述的信息采集装置， 其中， 所述复眼仿生单元为复眼仿生摄 像机或者曲面设置的第二图像传感器阵列。
6. 根据权利要求 1至 5中任一项所述的信息采集装置， 其中， 还包括控制单元， 分别与所述背单眼仿生单元、 所述侧单眼仿生单元和所述复眼仿生单元相连， 用于接收各单元的信息。
7. 一种身份识别系统， 其中， 包括如权利要求 1至 6中任一项所述的用于采集物 理身份信息的信息采集装置， 以及用于对所述物理身份信息进行识别的信息识 别装置。
8. 根据权利要求 7所述的身份识别系统， 其中， 所述信息识别装置包括： 识别服 务器、 信息数据库； 所述信息数据库用于存储数字身份信息； 所述识别服务器 用于将所述物理身份信息在所述信息数据库存储的数字身份信息中进行身份识 另 |J。
9. 一种信息采集方法， 包括以下处理过程：
通过背单眼仿生单元采集环境参数信息；
通过侧单眼仿生单元采集目标轮廓信息；
根据所述环境参数信息和目标轮廓信息， 并通过复眼仿生单元采集动态图 像信息。
10. 一种身份识别方法， 包括如权利要求 9所述的信息采集方法采集物理身份信息 的处理步骤， 还包括根据所述物理身份信息进行识别的处理步骤。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110371545.XA CN103123690B (zh) | 2011-11-21 | 2011-11-21 | 信息采集装置和方法以及身份识别系统和方法 |
| CN201110371545.X | 2011-11-21 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2013075497A1 true WO2013075497A1 (zh) | 2013-05-30 |
Family
ID=48454663
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2012/077957 WO2013075497A1 (zh) | 2011-11-21 | 2012-06-29 | 信息采集装置和方法以及身份识别系统和方法 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103123690B (zh) |
| WO (1) | WO2013075497A1 (zh) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104135612B (zh) * | 2014-07-11 | 2018-05-01 | 努比亚技术有限公司 | 一种可调节拍摄主体位置的拍摄方法和拍摄装置 |
| CN104270242B (zh) * | 2014-09-27 | 2017-12-19 | 杭州电子科技大学 | 一种用于网络数据加密传输的加解密装置 |
| CN107444341A (zh) * | 2017-07-25 | 2017-12-08 | 北京新能源汽车股份有限公司 | 一种汽车控制方法、装置及汽车 |
| CN108960072B (zh) * | 2018-06-06 | 2022-05-10 | 华为技术有限公司 | 一种步态识别的方法及设备 |
| CN109886166A (zh) * | 2019-01-31 | 2019-06-14 | 杭州创匠信息科技有限公司 | 基于偏振特性的防伪方法和装置 |
| CN109886196A (zh) * | 2019-02-21 | 2019-06-14 | 中水北方勘测设计研究有限责任公司 | 基于bim加gis视频监控的人员轨迹追溯系统及方法 |
| CN109902611B (zh) * | 2019-02-22 | 2021-03-02 | 矩阵元技术（深圳）有限公司 | 目标证件的检测方法、装置和终端设备 |
| CN113052197B (zh) * | 2019-12-28 | 2024-03-12 | 中移(成都)信息通信科技有限公司 | 身份识别的方法、装置、设备和介质 |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050110672A1 (en) * | 2003-10-10 | 2005-05-26 | L-3 Communications Security And Detection Systems, Inc. | Mmw contraband screening system |
| CN1932841A (zh) * | 2005-10-28 | 2007-03-21 | 南京航空航天大学 | 基于仿生复眼的运动目标检测装置及其方法 |
| CN101201955A (zh) * | 2007-12-19 | 2008-06-18 | 中国科学院长春光学精密机械与物理研究所 | Atm的智能监控系统及其方法 |
| US20090309728A1 (en) * | 2007-03-16 | 2009-12-17 | Fujitsu Limited | Object detection method and object detection system |
| CN102097832A (zh) * | 2010-10-27 | 2011-06-15 | 国家电网公司 | 基于物联网的充换电监控系统及方法 |
| CN102157023A (zh) * | 2011-01-13 | 2011-08-17 | 中华人民共和国深圳出入境检验检疫局 | 出/入境旅客自助申报与监管信息系统 |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2483797Y (zh) * | 2001-03-16 | 2002-03-27 | 王广生 | 可控视野显示装置 |
-
2011
- 2011-11-21 CN CN201110371545.XA patent/CN103123690B/zh active Active
-
2012
- 2012-06-29 WO PCT/CN2012/077957 patent/WO2013075497A1/zh active Application Filing
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050110672A1 (en) * | 2003-10-10 | 2005-05-26 | L-3 Communications Security And Detection Systems, Inc. | Mmw contraband screening system |
| CN1932841A (zh) * | 2005-10-28 | 2007-03-21 | 南京航空航天大学 | 基于仿生复眼的运动目标检测装置及其方法 |
| US20090309728A1 (en) * | 2007-03-16 | 2009-12-17 | Fujitsu Limited | Object detection method and object detection system |
| CN101201955A (zh) * | 2007-12-19 | 2008-06-18 | 中国科学院长春光学精密机械与物理研究所 | Atm的智能监控系统及其方法 |
| CN102097832A (zh) * | 2010-10-27 | 2011-06-15 | 国家电网公司 | 基于物联网的充换电监控系统及方法 |
| CN102157023A (zh) * | 2011-01-13 | 2011-08-17 | 中华人民共和国深圳出入境检验检疫局 | 出/入境旅客自助申报与监管信息系统 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103123690B (zh) | 2017-02-22 |
| CN103123690A (zh) | 2013-05-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2013075497A1 (zh) | 信息采集装置和方法以及身份识别系统和方法 | |
| US11321983B2 (en) | System and method for identifying and verifying one or more individuals using facial recognition | |
| US9971920B2 (en) | Spoof detection for biometric authentication | |
| CN109461003A (zh) | 基于多视角的多人脸场景刷脸支付风险防控方法和设备 | |
| US20040218788A1 (en) | Three-dimensional ear biometrics system and method | |
| JP2006293644A (ja) | 情報処理装置、情報処理方法 | |
| CN107688775A (zh) | 基于电梯场景的双目摄像头人脸识别鉴权系统及方法 | |
| WO2010012215A1 (zh) | 一种双摄像头人脸识别装置和方法 | |
| Nixon et al. | On use of biometrics in forensics: gait and ear | |
| TW200905577A (en) | Iris recognition system | |
| CN108446690A (zh) | 一种基于多视角动态特征的人脸活体检测方法 | |
| JP2014529815A (ja) | バイオメトリクスタイプのアクセス制御システムにおける不正の検出 | |
| CN110309693A (zh) | 多层次状态侦测系统与方法 | |
| Hossain et al. | Next generation identity verification based on face-gait Biometrics | |
| CN109800638A (zh) | 一种基于人脸识别技术的重点人监控方法 | |
| WO2006053375A1 (en) | Computer-based method and system for identifying a potential partner | |
| CN110148234A (zh) | 校园刷脸接送交互方法、存储介质及系统 | |
| Rabie et al. | Nfc-based person-specific assisting system in home environment | |
| Yu et al. | Research on face anti-spoofing algorithm based on image fusion | |
| CN106778577A (zh) | 净水器使用者身份识别方法 | |
| CN106778578A (zh) | 净水器用户身份识别方法 | |
| Drosou et al. | Unobtrusive behavioral and activity-related multimodal biometrics: the ACTIBIO authentication concept | |
| Chowdhary | Analysis of Unimodal and Multimodal Biometric System | |
| Singh et al. | Adapted Facial Recognition And Spoofing Detection For Management Decision Making System: A Visually Impaired People Perspective | |
| Hemalatha et al. | A study of liveness detection in face biometric systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12852052 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 12852052 Country of ref document: EP Kind code of ref document: A1 |