WO2016201731A1 - 一种指纹识别方法、装置及电子设备 - Google Patents

一种指纹识别方法、装置及电子设备 Download PDF

Info

Publication number
WO2016201731A1
WO2016201731A1 PCT/CN2015/082900 CN2015082900W WO2016201731A1 WO 2016201731 A1 WO2016201731 A1 WO 2016201731A1 CN 2015082900 W CN2015082900 W CN 2015082900W WO 2016201731 A1 WO2016201731 A1 WO 2016201731A1
Authority
WO
WIPO (PCT)
Prior art keywords
fingerprint
feature
user
matching degree
matching
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.)
Ceased
Application number
PCT/CN2015/082900
Other languages
English (en)
French (fr)
Inventor
钟焰涛
傅文治
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yulong Computer Telecommunication Scientific Shenzhen Co Ltd
Original Assignee
Yulong Computer Telecommunication Scientific Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yulong Computer Telecommunication Scientific Shenzhen Co Ltd filed Critical Yulong Computer Telecommunication Scientific Shenzhen Co Ltd
Publication of WO2016201731A1 publication Critical patent/WO2016201731A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1365—Matching; Classification
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1347—Preprocessing; Feature extraction

Definitions

  • the invention belongs to the field of intelligent identification technology of biometrics, and in particular relates to a fingerprint identification method, device and electronic device.
  • Fingerprint recognition technology is the most widely used biometric recognition technology.
  • the technology realizes fingerprint recognition based on various feature points such as endpoints, bifurcation points, bifurcation points, isolated points, ring points and short lines contained in the fingerprint. Fingerprint recognition has been integrated into many high-end mobile phones.
  • Fingerprint recognition technology includes fingerprint feature extraction and fingerprint recognition and recognition.
  • the conventional fingerprint identification scheme stores, in the fingerprint feature database, a fingerprint feature used as a fingerprint template, which is pre-acquired (for example, can be collected when the user registers the fingerprint recognition function). Therefore, after extracting the fingerprint feature of the fingerprint input by the user, the extracted user fingerprint feature may be compared with each fingerprint template included in the fingerprint feature database, if the fingerprint feature of the user matches the feature points of a certain fingerprint template. When the set threshold is reached, the user is identified.
  • the traditional fingerprint identification scheme cannot effectively identify different fingerprints of the user, for example, when the fingerprint of the user is dry or peeled, some feature points are difficult to extract, and the extracted fingerprint features temporarily occur.
  • the traditional fingerprint identification scheme is easy to generate problems that cannot be effectively identified.
  • the object of the present invention is to provide a fingerprint identification method, device and electronic device, which aims to solve the problem that the traditional fingerprint identification scheme cannot effectively identify different states of the user's fingerprint, and enhance the fingerprint recognition. Accuracy.
  • a fingerprint identification method comprising:
  • the fingerprint feature database includes at least one fingerprint template, each of the fingerprints
  • the template includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponding to a user fingerprint state, and each set of feature point weights is used to calculate a matching degree value of the user fingerprint in the corresponding user fingerprint state, P is a natural number greater than one;
  • the fingerprint recognition is successful.
  • the first fingerprint feature and the second fingerprint feature respectively comprise a corresponding number of feature points
  • the feature points include an endpoint, a bifurcation point, a bifurcation point, an isolated point, a ring point, and a short Pattern.
  • the first fingerprint feature is matched and identified based on the preset fingerprint feature database, and the optimal matching degree of the first fingerprint feature is obtained, including:
  • the matching degree with the largest value is selected as the optimal matching degree of the first fingerprint feature.
  • the matching result includes an n-dimensional 0-1 vector (a 1 , a 2 , . . . , a n ), where n is a natural number greater than 1, wherein
  • a i 1 indicates that the i-th feature point in the first fingerprint feature matches the i-th feature point in the second fingerprint feature;
  • the above method preferably, further includes:
  • the set of feature point weights used in calculating the optimal matching degree is adjusted according to the matching result used when calculating the optimal matching degree.
  • the above method preferably, further includes:
  • the feature point weight training is performed on the fingerprint template in a preset P user fingerprint state, and the P group feature point weights corresponding to the P user fingerprint states are obtained.
  • the above method preferably, further includes:
  • fingerprint recognition fails.
  • a fingerprint identification device comprising:
  • a fingerprint acquiring module configured to acquire a fingerprint input by a user
  • a feature extraction module configured to perform feature point extraction on the fingerprint to obtain a first fingerprint feature to be identified
  • a matching identification module configured to perform matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and obtain an optimal matching degree of the first fingerprint feature; wherein the fingerprint feature database includes at least one fingerprint template
  • Each of the fingerprint templates includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponds to a user fingerprint state, and each set of feature point weights is used to calculate a user of the corresponding user fingerprint state.
  • the matching value of the fingerprint, P is a natural number greater than 1;
  • the first result determining module is configured to determine that the fingerprint recognition is successful when the optimal matching degree reaches a set threshold.
  • the matching identification module includes:
  • a matching unit configured to match the first fingerprint feature with a second fingerprint feature included in each fingerprint template in the fingerprint feature database to obtain a matching result
  • a calculating unit configured to perform weighting calculation on the matching result by using each set of feature point weights of the fingerprint template, to obtain a matching degree of the first fingerprint feature in various user fingerprint states
  • a selecting unit configured to select a matching degree with the largest value from the series of matching degrees of the first fingerprint feature corresponding to each fingerprint template and each user fingerprint state, and use the same as the optimal matching of the first fingerprint feature degree.
  • the above device preferably, further comprises:
  • a weight adjustment module configured to adjust a set of feature point weights used when calculating the optimal matching degree according to a matching result used when calculating the optimal matching degree when fingerprint identification is successful .
  • the above device preferably, further comprises:
  • a pre-processing module configured to advance the fingerprint template in a preset P user fingerprint state
  • the row feature point weight training is performed, and the P group feature point weights corresponding to the P user fingerprint states are obtained one by one.
  • the above device preferably, further comprises:
  • the second result determining module is configured to determine that the fingerprint recognition fails when the optimal matching degree does not reach the set threshold.
  • An electronic device includes a communication bus 1002, a transceiver, a memory, and a processor, wherein:
  • the communication bus 1002 is configured to implement connection communication between the transceiver device, the memory, and the processor;
  • the memory stores a set of program codes, and the processor calls the program code stored in the memory to perform the following operations:
  • the transceiver device is configured to acquire a fingerprint input by a user
  • the processor is configured to perform feature point extraction on the fingerprint to obtain a first fingerprint feature to be identified
  • the processor is further configured to perform matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and acquire an optimal matching degree of the first fingerprint feature; wherein the fingerprint feature database includes at least one a fingerprint template, each of the fingerprint templates includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponding to a user fingerprint state, and each set of feature point weights is used to calculate a corresponding user fingerprint state
  • the matching value of the user fingerprint, P is a natural number greater than 1;
  • the processor is further configured to determine that the fingerprint recognition succeeds when the optimal matching degree reaches a set threshold.
  • the processor performs matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and acquiring an optimal matching degree of the first fingerprint feature specifically includes:
  • the processor is further configured to match the first fingerprint feature with a second fingerprint feature included in each fingerprint template in the fingerprint feature database to obtain a matching result;
  • the processor is further configured to perform weighting calculation on the matching result by using each set of feature point weights of the fingerprint template, to obtain a matching degree of the first fingerprint feature in various user fingerprint states;
  • the processor is further configured to select a matching degree with the largest value from the series of matching degrees of the first fingerprint feature corresponding to each fingerprint template and each user fingerprint state, and use the first fingerprint feature as the first fingerprint feature. Optimal matching.
  • the processor is further configured to perform the following steps:
  • the set of feature point weights used when calculating the optimal matching degree is adjusted.
  • the processor is further configured to perform the following steps:
  • the feature point weight training is performed on the fingerprint template in a preset P user fingerprint state, and the P group feature point weights corresponding to the P user fingerprint states are obtained.
  • the processor is further configured to perform the following steps:
  • the judgment recognition result is that the fingerprint recognition fails.
  • the present invention performs matching matching on the fingerprint features of the user to be identified based on the preset fingerprint feature database, and obtains the optimal matching degree of the fingerprint feature of the user, and then reaches the set threshold in the optimal matching degree.
  • the fingerprint feature database includes at least one fingerprint template, where the fingerprint template includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponds to a user fingerprint state, and each set of feature point weights is used. Calculate the matching value of the user's fingerprint in the corresponding state.
  • the present invention matches multiple sets of different feature point weights for each fingerprint template in the database, so that when the fingerprint of the user changes due to factors such as drying or peeling, the extracted fingerprint features change.
  • the feature point weight group of the appropriate state By using the feature point weight group of the appropriate state, a more accurate matching degree is calculated, the problem of the traditional scheme is solved, and the accuracy of fingerprint recognition is improved.
  • Embodiment 1 is a flowchart of Embodiment 1 of a fingerprint identification method provided by the present application;
  • Embodiment 2 is a flowchart of Embodiment 2 of a fingerprint identification method provided by the present application;
  • Embodiment 3 is a schematic diagram of a weight group training process provided by Embodiment 2 of the present application.
  • Embodiment 4 is a flowchart of Embodiment 3 of a fingerprint identification method provided by the present application.
  • FIG. 5 is a flowchart of Embodiment 4 of a fingerprint identification method provided by the present application.
  • FIG. 6 is a schematic structural diagram of a fingerprint identification apparatus according to Embodiment 5 of the present application.
  • FIG. 10 is a schematic structural diagram of an electronic device according to Embodiment 6 of the present application.
  • a first embodiment of the present invention discloses a fingerprint identification method, which can be applied to an electronic device such as a smart phone or a tablet computer, for example, specifically applicable to user identity authentication of a corresponding service of an electronic device.
  • the method can include the following steps:
  • S101 Acquire a fingerprint input by a user.
  • the fingerprint input by the user through the fingerprint sensor can be obtained.
  • S102 Perform feature point extraction on the fingerprint to obtain a first fingerprint feature to be identified.
  • the fingerprint texture is not continuous, smooth and straight, but features such as interruption, bifurcation or turning.
  • the feature points generated by these features provide confirmation information about the uniqueness of the fingerprint.
  • the most typical feature points are the endpoint and Bifurcation points, other feature points also include divergence points, isolated points, ring points, short lines, etc.
  • the parameters of the feature points include direction, curvature, and position.
  • user fingerprint features including various feature points such as an endpoint, a bifurcation point, a bifurcation point, an isolated point, a ring point, and a short pattern can be obtained.
  • S103 Matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and acquiring an optimal matching degree of the first fingerprint feature; wherein the fingerprint feature database includes at least one fingerprint template, where each The fingerprint template includes a second fingerprint feature and a P group feature point weight.
  • Each set of feature point weights corresponds to a user fingerprint state, and each set of feature point weights is used to calculate the matching degree of the user fingerprints in the corresponding user fingerprint state. For the value, P is a natural number greater than one.
  • the user's fingerprints tend to be too dry, too moist or peeling, which makes it difficult to extract certain fingerprint features, which affects the accuracy of fingerprint recognition. Based on this, the user first divides the user's fingerprint status into normal. Dry, moist and peeling four states.
  • each fingerprint template in the fingerprint feature database by learning the extraction and recognition of fingerprint features in various states, four sets of feature point weights are assigned to each fingerprint template in the fingerprint feature database, and each set of feature point weights corresponds to a corresponding user fingerprint state.
  • the fingerprint feature contained in the fingerprint template can be represented as an n-dimensional vector containing n eigenvalues: (c 1 , c 2 , . . . , c n ),
  • the feature value c i is a specific representation of the i-th feature point in the fingerprint template; and corresponding to the normal, dry, wet and peeled state of the user fingerprint, the fingerprint template includes four sets of feature point weights: (w 11 ; w 12 , ..., w 1n ), (w 21 , w 22 , ..., w 2n ), (w 31 , w 32 , ..., w 3n ) and (w 41 , w 42 , ..., w 4n ).
  • the invention increases the weight of the feature points that are easy to be extracted and recognized in the corresponding fingerprint state, and reduces the weight of the feature points that are difficult to be extracted and recognized, thereby realizing the matching of the fingerprint features in the state.
  • the user fingerprint feature is sequentially matched with each fingerprint template included in the fingerprint feature database. Specifically, in the process of matching with each fingerprint template, first, matching each feature point included in the user fingerprint feature with each feature point included in the fingerprint template, an n-dimensional 0-1 can be obtained.
  • the matching result represented by the vector (a 1 , a 2 , ..., a n ), wherein
  • the matching degree values in the four fingerprint states of normal, dry, wet, and peeling can be obtained separately: p 1 , p 2 , p 3 , and p 4 .
  • the matching degree calculated for each fingerprint state under each fingerprint template is integrated, and the matching degree with the largest value is selected as the user fingerprint feature.
  • the optimal matching degree wherein the fingerprint state corresponding to the optimal matching degree is the most suitable state corresponding to the user fingerprint, that is, the fingerprint corresponding to the optimal matching degree
  • the status reflects the actual status of the user's fingerprint.
  • the step specifically compares the optimal matching degree of the user fingerprint feature with a preset matching threshold. If the optimal matching degree reaches the set threshold, the user fingerprint matching is successfully performed. Thus, the user can be identified.
  • the threshold t can be set to a real number satisfying 0 ⁇ t ⁇ 1. In the actual recognition scenario, the t value is too high to cause unrecognized, so the threshold t is generally set to a value less than 0.2.
  • the division of the fingerprint state of the user, and the feature point weight of the corresponding group number given to the fingerprint template on the basis of the present application are only exemplary descriptions of the solution of the present application. Personnel can divide or set the user fingerprint status and the number of feature point weights according to the various states in which the user's fingerprints appear in real life.
  • the present invention performs matching matching on the fingerprint features of the user to be identified based on the preset fingerprint feature database, and obtains the optimal matching degree of the fingerprint feature of the user, and then reaches the set threshold in the optimal matching degree.
  • the fingerprint feature database includes at least one fingerprint template, where the fingerprint template includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponds to a user fingerprint state, and each set of feature point weights is used. Calculate the matching value of the user's fingerprint in the corresponding state.
  • the present invention matches multiple sets of different feature point weights for each fingerprint template in the database, so that when the fingerprint of the user changes due to factors such as drying or peeling, the extracted fingerprint features change.
  • the feature point weight group of the appropriate state By using the feature point weight group of the appropriate state, a more accurate matching degree is calculated, the problem of the traditional scheme is solved, and the accuracy of fingerprint recognition is improved.
  • the fingerprint identification method may include the following preprocessing process:
  • S101' Perform feature point weight training on the fingerprint template in a preset P user fingerprint state, and obtain P group feature point weights corresponding to the P user fingerprint states one by one.
  • the feature point weight of the corresponding group number is given in advance according to the number of user fingerprint states.
  • the feature point weight group corresponding to the state is trained by learning the fingerprint feature extraction and recognition process in each fingerprint state, and each group weight The acquisition of values requires separate training procedures.
  • the training process of the weight group may include the following steps:
  • the fingerprint template contains a total of n feature points, denoted as (c 1 , c 2 , . . . , c n ).
  • Group (1/n, 1/n, ..., 1/n) and set the matching threshold t to a real number satisfying 0 ⁇ t ⁇ 1, because the value of t is too high, it is easy to cause no recognition, so
  • This embodiment specifically sets t to a value less than 0.2.
  • S302 Acquire a user fingerprint and extract a user fingerprint feature.
  • the fingerprint input by the user through the fingerprint sensor in the dry state of the fingerprint is obtained, and the fingerprint feature is extracted.
  • step S303 Identify the fingerprint feature of the user, and determine whether the recognition is successful. If the recognition is successful, step S304 is performed, otherwise, the process proceeds to step S305.
  • the fingerprint template, the weight group and the set threshold are used to identify the fingerprint feature of the user.
  • the specific identification process refer to the description of the first embodiment, which is not described in detail herein.
  • the weights of the corresponding feature points in the weight group are adjusted, and the calculation formula used for the adjustment is as follows:
  • w i ' is the modified and adjusted feature point weight
  • w i is the feature point weight used for matching and identifying the user fingerprint
  • v i is an intermediate value in the process of calculating w i '
  • the effect is to maintain the adjusted sum of the weights as 1, ie
  • the invention increases the weight of the feature points that are easy to be extracted and recognized in the corresponding fingerprint state, reduces the weight of the feature points that are difficult to be extracted and recognized, and realizes the adjustment of the matching degree of the fingerprint features in the state, and ensures the state. A more accurate match value can be obtained.
  • SS305 Determine whether the training process is sufficient, if it is enough, end the training process; otherwise, go to step S302 to enter the next round of training.
  • the invention continuously performs the iterative modification process of the feature point weights, and after sufficient training, finally obtains the feature point weight group which is required to be corresponding to the state and has a reasonable weight distribution.
  • the present invention continues to perform the multiple fingerprint training in different states to adjust the weight of the fingerprint template, and can form a precise identification of various state fingerprints after convergence. Multiple sets of weights.
  • the fingerprint identification method may further include the following steps:
  • S105 If the fingerprint identification is successful, adjusting a set of feature point weights used when calculating the optimal matching degree according to the matching result used when calculating the optimal matching degree.
  • the feature matching situation according to the successful recognition may also be used according to the feature matching in the successful recognition.
  • the weight of the fingerprint template is adjusted to adapt to the small change of the fingerprint.
  • the fingerprint identification method may further include the following steps:
  • the optimal matching degree corresponding to the user fingerprint feature is small, and the threshold value is not reached, the matching between the fingerprint of the user and the fingerprint template in the database is poor, so that the fingerprint identification fails and the user is not recognized.
  • the fifth embodiment discloses a fingerprint identification device, which is disclosed in the first embodiment to the fourth embodiment.
  • the fingerprint identification method corresponds.
  • the device includes a fingerprint acquiring module 100, a feature extraction module 200, a matching identification module 300, and a first result decision module 400.
  • the fingerprint obtaining module 100 is configured to acquire a fingerprint input by a user.
  • the feature extraction module 200 is configured to perform feature point extraction on the fingerprint to obtain a first fingerprint feature to be identified.
  • the matching identification module 300 is configured to perform matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and acquire an optimal matching degree of the first fingerprint feature; wherein the fingerprint feature database includes at least one fingerprint a template, each of the fingerprint templates includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponding to a user fingerprint state, and each set of feature point weights is used to calculate a corresponding user fingerprint state
  • the matching value of the user's fingerprint, P is a natural number greater than 1.
  • the matching identification module 300 includes a matching unit, a computing unit, and
  • a matching unit configured to match the first fingerprint feature with a second fingerprint feature included in each fingerprint template in the fingerprint feature database to obtain a matching result and a selection unit.
  • a calculating unit configured to perform weighting calculation on the matching result by using each set of feature point weights of the fingerprint template, to obtain a matching degree of the first fingerprint feature in various user fingerprint states
  • a selecting unit configured to select a matching degree with the largest value from the series of matching degrees of the first fingerprint feature corresponding to each fingerprint template and each user fingerprint state, and use the same as the optimal matching of the first fingerprint feature degree.
  • the first result decision module 400 is configured to determine that the fingerprint recognition is successful when the optimal matching degree reaches a set threshold.
  • the device further includes a pre-processing module 500, configured to perform feature point weight training on the fingerprint template in advance in a preset P user fingerprint state, and obtain the P group feature point weights corresponding to P user fingerprint states.
  • a pre-processing module 500 configured to perform feature point weight training on the fingerprint template in advance in a preset P user fingerprint state, and obtain the P group feature point weights corresponding to P user fingerprint states.
  • the device further includes a weight adjustment module 600, configured to calculate the matching result when the fingerprint is successfully determined according to the matching result used when calculating the optimal matching degree.
  • the set of feature point weights used in the optimal matching degree are adjusted.
  • the apparatus further includes a second result decision module 700, configured to determine that the recognition result is a fingerprint recognition failure when the optimal matching degree does not reach the set threshold.
  • the embodiment discloses an electronic device, the electronic device at least one transceiver device 1003, at least one processor 1001, such as a CPU, a memory 1004 and at least one communication bus 1002.
  • the communication bus 1002 is used to connect the transceiver 1003, the processor 1001, and the memory 1004.
  • the above memory 1004 may be a high speed RAM memory or a non-volatile memory such as a disk memory.
  • the above-mentioned memory 1004 is further configured to store a set of program codes, and the transceiver device 1003 and the processor 1001 are configured to call the program code stored in the memory 1004, and perform the following operations:
  • the transceiver device 1003 is configured to acquire a fingerprint input by a user
  • the processor 1001 is configured to perform feature point extraction on the fingerprint to obtain a first fingerprint feature to be identified;
  • the processor 1001 is further configured to perform matching and identifying the first fingerprint feature based on a preset fingerprint feature database, and acquire an optimal matching degree of the first fingerprint feature; wherein the fingerprint feature database includes at least a fingerprint template, each of the fingerprint templates includes a second fingerprint feature and a P group feature point weight, each set of feature point weights corresponding to a user fingerprint state, and each set of feature point weights is used to calculate a corresponding user fingerprint
  • the matching value of the user's fingerprint in the state, P is a natural number greater than 1;
  • the processor 1001 is further configured to: when the optimal matching degree reaches a set threshold, determine that the recognition result is successful fingerprint identification.
  • the processor 1001 performs the matching and identifying the first fingerprint feature based on the preset fingerprint feature database, and acquiring the optimal matching degree of the first fingerprint feature specifically includes:
  • the processor 1001 is further configured to match the first fingerprint feature with a second fingerprint feature included in each fingerprint template in the fingerprint feature database to obtain a matching result;
  • the processor 1001 is further configured to perform weighting calculation on the matching result by using each set of feature point weights of the fingerprint template, to obtain a matching degree of the first fingerprint feature in various user fingerprint states;
  • the processor 1001 is further configured to correspond to each fingerprint template and each user from the first fingerprint feature. Among the series of matching degrees of the fingerprint state, the matching degree with the largest value is selected as the optimal matching degree of the first fingerprint feature.
  • the processor 1001 is further configured to perform the following steps:
  • the set of feature point weights used when calculating the optimal matching degree is adjusted.
  • the processor 1001 is further configured to perform the following steps:
  • the feature point weight training is performed on the fingerprint template in a preset P user fingerprint state, and the P group feature point weights corresponding to the P user fingerprint states are obtained.
  • the processor 1001 is further configured to perform the following steps:
  • the judgment recognition result is that the fingerprint recognition fails.
  • the electronic device of the embodiment can accurately identify the user's fingerprint under various states (such as normal, too dry, too humid, peeling, etc.), and does not change due to the state of the user's fingerprint.
  • states such as normal, too dry, too humid, peeling, etc.
  • the problem of difficult identification is high, and the recognition accuracy is high, which improves the user experience.
  • the present application can be implemented by means of software plus a necessary general hardware platform. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product in essence or in the form of a software product, which may be stored in a storage medium such as a ROM/RAM or a disk. , an optical disk, etc., includes instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to perform the methods described in various embodiments of the present application or portions of the embodiments.
  • a computer device which may be a personal computer, server, or network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Security & Cryptography (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Collating Specific Patterns (AREA)
  • Image Analysis (AREA)

Abstract

一种指纹识别方法、装置及电子设备，其基于预设的指纹特征数据库对待识别的用户指纹特征进行匹配识别，并获取所述用户指纹特征的最优匹配度，之后在所述最优匹配度达到设定的阈值时，识别该用户。其中，所述指纹特征数据库包含至少一个指纹模板，该指纹模板包含P组特征点权值，每组权值对应一用户指纹状态，且每组特征点权值用于计算其对应状态下用户指纹的匹配度数值。针对用户指纹的不同状态，预先为数据库中的每一指纹模板匹配了多组不同的特征点权值，从而当用户指纹因干燥或脱皮等因素，导致提取的指纹特征发生变化时，可通过采用合适状态的特征点权值组，来计算出一较为准确的匹配度，提升了指纹识别的准确度。

Description

一种指纹识别方法、装置及电子设备
本申请要求于2015年6月17日提交中国专利局、申请号为201510337023.6，发明名称为“一种指纹识别方法、装置及电子设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明属于生物特征的智能识别技术领域，尤其涉及一种指纹识别方法、装置及电子设备。
背景技术
指纹识别技术是应用最广泛的一种生物特征识别技术，该技术基于指纹中包含的终结点、分叉点、分歧点、孤立点、环点、短纹等各种特征点实现指纹识别。目前许多高端手机中已经集成了指纹识别功能。
指纹识别技术包含指纹特征提取，以及指纹特征的匹配识别两个阶段。传统的指纹识别方案，在指纹特征数据库中存储有预先采集(例如，可在用户注册指纹识别功能时采集)的、用于作为指纹模板的指纹特征。从而在提取用户所输入指纹的指纹特征后，可将提取的用户指纹特征与指纹特征数据库中包含的各指纹模板进行比对，如果用户的指纹特征与某一指纹模板中各特征点的匹配度达到设定的阈值，则识别该用户。
然而，传统的指纹识别方案无法针对用户指纹的不同状态，对其进行有效识别，例如，当用户指纹因干燥或脱皮等因素，而导致某些特征点难以提取，进而导致提取的指纹特征暂时发生变化时，利用传统的指纹识别方案易产生无法有效识别的问题。
发明内容
有鉴于此，本发明的目的在于提供一种指纹识别方法、装置及电子设备，旨在解决传统的指纹识别方案无法针对用户指纹的不同状态，对其进行有效识别这一问题，提升指纹识别的准确度。
为此，本发明公开如下技术方案：
一种指纹识别方法，包括：
获取用户输入的指纹；
对所述指纹进行特征点提取，得到待识别的第一指纹特征；
基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
如果所述最优匹配度达到设定的阈值，则指纹识别成功。
上述方法，优选的，所述第一指纹特征、所述第二指纹特征分别包含相应个数的特征点，所述特征点包括终结点、分叉点、分歧点、孤立点、环点、短纹。
上述方法，优选的，所述基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度，包括：
将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
上述方法，优选的，所述匹配结果包括n维0-1向量(a1,a2,......,an)，n为大于1的自然数，其中，
ai＝1表示所述第一指纹特征中的第i个特征点，与所述第二指纹特征中的第i个特征点相匹配；
ai＝0表示所述第一指纹特征中的第i个特征点，与所述第二指纹特征中的第i个特征点不匹配，i＝1,2,......,n。
上述方法，优选的，还包括：
如果指纹识别成功，则依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
上述方法，优选的，还包括：
预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
上述方法，优选的，还包括：
如果所述最优匹配度未达到设定的阈值，则指纹识别失败。
一种指纹识别装置，包括：
指纹获取模块，用于获取用户输入的指纹；
特征提取模块，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征；
匹配识别模块，用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
第一结果判决模块，用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
上述装置，优选的，所述匹配识别模块包括：
匹配单元，用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
计算单元，用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
选取单元，用于从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
上述装置，优选的，还包括：
权值调整模块，用于在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
上述装置，优选的，还包括：
预处理模块，用于预先在预设的P种用户指纹状态下，对所述指纹模板进 行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
上述装置，优选的，还包括：
第二结果判决模块，用于在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
一种电子设备，包括通信总线1002、收发装置、存储器以及处理器，其中：
所述通信总线1002，用于实现所述收发装置、所述存储器以及所述处理器之间的连接通信；
所述存储器中存储一组程序代码，且处理器调用存储器中存储的程序代码，用于执行以下操作：
所述收发装置，用于获取用户输入的指纹；
所述处理器，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征；
所述处理器，还用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
所述处理器，还用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
上述电子设备，优选的，所述处理器基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度具体包括：
所述处理器，还用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
所述处理器，还用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
所述处理器，还用于从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
上述电子设备，优选的，所述处理器还用于执行如下步骤：
在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
上述电子设备，优选的，所述处理器还用于执行如下步骤：
预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
上述电子设备，优选的，所述处理器还用于执行如下步骤：
在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
由以上方案可知，本发明基于预设的指纹特征数据库对待识别的用户指纹特征进行匹配识别，并获取所述用户指纹特征的最优匹配度，之后在所述最优匹配度达到设定的阈值时，识别该用户。其中，所述指纹特征数据库包含至少一个指纹模板，该指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应状态下用户指纹的匹配度数值。可见，本发明预先针对用户指纹的不同状态，为数据库中的每一指纹模板匹配了多组不同的特征点权值，从而当用户指纹因干燥或脱皮等因素，导致提取的指纹特征发生变化时，可通过采用合适状态的特征点权值组，来计算出一较为准确的匹配度，解决了传统方案存在的问题，提升了指纹识别的准确度。
附图说明
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，下面描述中的附图仅仅是本发明的实施例，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，还可以根据提供的附图获得其他的附图。
图1是本申请提供的指纹识别方法实施例一的流程图；
图2是本申请提供的指纹识别方法实施例二的流程图；
图3是本申请实施例二提供的权值组训练过程示意图；
图4是本申请提供的指纹识别方法实施例三的流程图；
图5是本申请提供的指纹识别方法实施例四的流程图；
图6-图9是本申请实施例五提供的指纹识别装置的结构示意图；
图10是本申请实施例六提供的电子设备的结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
实施例一
本发明实施例一公开一种指纹识别方法，该指纹识别方法可应用于智能手机、平板电脑等电子设备中，例如，具体可应用于电子设备相应服务的用户身份认证中。参考图1，所述方法可以包括以下步骤：
S101：获取用户输入的指纹。
具体可获取用户通过指纹传感器所输入的指纹。
S102：对所述指纹进行特征点提取，得到待识别的第一指纹特征。
指纹纹路并不是连续的、平滑笔直的，而是经常出现中断、分叉或转折等特征，由这些特征产生的特征点提供了指纹唯一性的确认信息，其中最典型的特征点是终结点和分叉点，其他特征点还包括分歧点、孤立点、环点、短纹等，特征点的参数包括方向、曲率、位置。
从而在对用户输入的指纹进行特征点提取后，可得到包含终结点、分叉点、分歧点、孤立点、环点、短纹等各种特征点的用户指纹特征。
S103：基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数。
日常生活中，用户指纹往往会出现过于干燥、过于湿润或脱皮等状况，导致某些指纹特征难以被提取，进而影响指纹识别的准确度，基于此，本实施例首先将用户指纹状态划分为正常、干燥、湿润及脱皮四种状态。
与此同时，通过学习各种状态下指纹特征的提取、识别情况，为指纹特征数据库中的各指纹模板赋予4组特征点权值，每组特征点权值对应一相应的用户指纹状态。
假设一个指纹模板中共包含n个特征点，则指纹模板包含的指纹特征可表示为一包含n个特征值的n维向量：(c1,c2,......,cn)，其中，特征值ci为指纹模板中第i个特征点的具体表示；且对应于用户指纹的正常、干燥、湿润及脱皮4种状态，指纹模板包含4组特征点权值：(w11，w12，......，w1n)、(w21，w22，......，w2n)、(w31，w32，......，w3n)及(w41，w42，......，w4n)。
特征点权值wji越大，表征其对应的特征点在第i种状态下越易被提取及识别，否则，特征点权值wji越小，表征其对应的特征点在第i种状态下越难易被提取及识别，即本发明通过增加相应指纹状态下易被提取、识别的特征点的权值，降低难以被提取、识别的特征点的权值，实现对该状态下指纹特征的匹配度数值进行调节，确保一较准确的匹配度数值，其中，j＝1,2,3,4；i＝1,2,......,n。
在对用户输入的指纹进行特征点提取，得到由多个特征点构成的待识别的用户指纹特征后，将用户指纹特征依次与指纹特征数据库中包含的每个指纹模板进行匹配。具体地，在与每一指纹模板进行匹配的过程中，首先，将用户指纹特征中包含的各个特征点与该指纹模板中包含的各特征点进行匹配，则可得到一采用n维0-1向量(a1，a2，......，an)表示的匹配结果，其中，
Figure PCTCN2015082900-appb-000001
在此基础上，利用以下公式计算用户指纹特征对应于每种指纹状态下的匹配度：
Figure PCTCN2015082900-appb-000002
从而，可分别得到正常、干燥、湿润及脱皮4种指纹状态下的匹配度数值：p1、p2、p3、p4。
在计算出用户指纹特征对应于每一指纹模板中各指纹状态的匹配度之后，综合各指纹模板下，针对各指纹状态计算出的匹配度，并从中选取出数值最大的匹配度作为用户指纹特征的最优匹配度，其中，最优匹配度所对应的指纹状态，即为用户指纹所对应的最合适状态，也就是说，最优匹配度所对应的指纹 状态反映了用户指纹的实际状态。
S104：如果所述最优匹配度达到设定的阈值，则指纹识别成功。
在以上各步骤的基础上，本步骤具体将用户指纹特征的最优匹配度与预先设定的匹配阈值进行比对，如果所述最优匹配度达到设定的阈值，则表征用户指纹匹配成功，从而，可识别该用户。
其中，阈值t可设置为一满足0＜t＜1的实数，实际识别场景中，t值过高易导致不识别，因此一般将阈值t设置为一小于0.2的数值。
需要说明的是，本申请对用户指纹状态的划分，以及在此基础上为指纹模板赋予的相应组数的特征点权值，仅为本申请方案的示例性说明，实际应用本申请时，技术人员可以依据现实生活中用户指纹较常出现的各种状态，对用户指纹状态及特征点权值的组数进行自行划分或设定。
由以上方案可知，本发明基于预设的指纹特征数据库对待识别的用户指纹特征进行匹配识别，并获取所述用户指纹特征的最优匹配度，之后在所述最优匹配度达到设定的阈值时，识别该用户。其中，所述指纹特征数据库包含至少一个指纹模板，该指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应状态下用户指纹的匹配度数值。可见，本发明预先针对用户指纹的不同状态，为数据库中的每一指纹模板匹配了多组不同的特征点权值，从而当用户指纹因干燥或脱皮等因素，导致提取的指纹特征发生变化时，可通过采用合适状态的特征点权值组，来计算出一较为准确的匹配度，解决了传统方案存在的问题，提升了指纹识别的准确度。
实施例二
本实施例具体对获取不同状态下的多组特征点权值这一预处理过程进行说明。如图2所示，所述指纹识别方法可以包括以下的预处理过程：
S101’：预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
针对指纹特征数据库中的每个指纹模板，都需预先依据用户指纹状态的数目为其赋予相应组数的特征点权值。本实施例具体通过在每种指纹状态下学习指纹特征的提取、识别过程，来训练出该状态所对应的特征点权值组，每组权 值的获取均需单独执行相应的训练过程。
下文以获取某指纹模板在干燥状态下对应的权值组为例，对权值组的训练过程进行描述。参考图3，权值组的训练过程可以包括以下步骤：
S301：设置初始权值和阈值。
假设该指纹模板共包含n个特征点，表示为(c1,c2,......,cn)，首先设置各特征点的初始权值为1/n，从而得到初始权值组(1/n,1/n,......,1/n)，同时设定匹配阈值t为一满足0＜t＜1的实数，由于t值过高容易导致不识别，因此本实施例具体将t设置为一小于0.2的数值。
S302：获取用户指纹，并提取用户指纹特征。
接下来，获取指纹干燥状态下用户通过指纹传感器输入的指纹，并提取其指纹特征。
S303：识别用户指纹特征，并判断是否识别成功。如果识别成功，则执行步骤S304，否则，转入步骤S305。
在此基础上，利用上述指纹模板、权值组及设定的阈值，对用户的指纹特征进行识别，具体识别过程可参考实施例一的描述，此处不再详述。
S304：修改权值。
如果识别成功，则依据用户指纹特征与指纹模板的匹配情况，对权值组中的相应特征点权值进行调整，实现调整所采用的计算公式如下：
Figure PCTCN2015082900-appb-000003
wi'＝δ·vi   (3)
Figure PCTCN2015082900-appb-000004
其中，wi'为修改、调整后的特征点权值，wi为本次匹配、识别用户指纹时所采用的特征点权值，vi为计算wi'过程中的一中间值；δ为归一化因子，其作用在于保持调整后各权值之和为1，即
Figure PCTCN2015082900-appb-000005
由上述训练过程可知，如果用户指纹的第i个特征点与指纹模板中的特征 值ci相匹配，则其对应的权值增大，否则减小，同时所有权值始终保持在区间(0，1)之间，且各权值之和为1。即本发明通过增加相应指纹状态下易被提取、识别的特征点的权值，降低难以被提取、识别的特征点的权值，实现对该状态下指纹特征的匹配度进行调节，确保该状态下可获取一较准确的匹配度数值。
SS305：判断训练过程是否足够，如果已足够，则结束训练过程；否则转至执行步骤S302，进入下一轮训练。
本发明通过对特征点权值迭代修改过程的不断执行，在进行充分训练后，最终可得出该状态对应所需的、权值分配较为合理的特征点权值组。
从而，预处理阶段中，本发明在提取指纹特征形成指纹模板后，继续通过不同状态下的多次指纹训练，实现对指纹模板的权值进行调整，收敛以后能够形成精确识别各种状态指纹的多组权值。
实施例三
本实施例三中，参考图4，所述指纹识别方法还可以包括以下步骤：
S105：如果指纹识别成功，则依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
即具体地，在采用指纹模板及其对应的多组特征点权值进行指纹识别的过程中，如果用户指纹识别成功，则还可以根据成功识别时的特征匹配情况，对成功识别时对应采用的指纹模板权值组进行权值调整，以适应指纹的微小变化，具体调整过程及采用的计算式可参考实施例二的描述。
实施例四
本实施例三中，参考图5，所述指纹识别方法还可以包括以下步骤：
S106：如果所述最优匹配度未达到设定的阈值，则指纹识别失败。
在用户指纹特征对应的最优匹配度较小，未达到设定的阈值时，表征用户指纹与数据库中指纹模板的匹配情况较差，从而指纹识别失败，不识别该用户。
实施例五
本实施例五公开一种指纹识别装置，该装置与实施例一至实施例四公开的 指纹识别方法相对应。
相应于实施例一，参考图6，所述装置包括指纹获取模块100，特征提取模块200、匹配识别模块300及第一结果判决模块400。
指纹获取模块100，用于获取用户输入的指纹。
特征提取模块200，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征。
匹配识别模块300，用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数。
其中，所述匹配识别模块300包括匹配单元、计算单元、
匹配单元，用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果和选取单元。
计算单元，用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
选取单元，用于从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
第一结果判决模块400，用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
相应于实施例二，参考图7，所述装置还包括预处理模块500，用于预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
相应于实施例三，参考图8，所述装置还包括权值调整模块600，用于在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
相应于实施例三，参考图9，所述装置还包括第二结果判决模块700，用于在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
对于本发明实施例五公开的指纹识别装置而言，由于其与实施例一至实施 例四公开的指纹识别方法相对应，所以描述的比较简单，相关相似之处请参见实施例一至实施例四中指纹识别方法部分的说明即可，此处不再详述。
实施例六
本实施例公开一种电子设备，所述电子设备至少一个收发装置1003，至少一个处理器1001，例如CPU，存储器1004和至少一个通信总线1002。
其中，上述通信总线1002用于连接上述收发装置1003、处理器1001和存储器1004。
上述存储器1004可以是高速RAM存储器，也可为非不稳定的存储器(non-volatile memory)，例如磁盘存储器。上述存储器1004还用于存储一组程序代码，上述收发装置1003和处理器1001用于调用存储器1004中存储的程序代码，执行如下操作：
所述收发装置1003，用于获取用户输入的指纹；
所述处理器1001，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征；
所述处理器1001，还用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
所述处理器1001，还用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
优选的，所述处理器1001基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度具体包括：
所述处理器1001，还用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
所述处理器1001，还用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
所述处理器1001，还用于从所述第一指纹特征对应于各指纹模板、各用户 指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
优选的，所述处理器1001还用于执行如下步骤：
在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
优选的，所述处理器1001还用于执行如下步骤：
预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
优选的，所述处理器1001还用于执行如下步骤：
在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
通过所述指纹识别装置，本实施例的电子设备可较为准确地识别各种状态(如正常、过于干燥、过于湿润、脱皮等状态)下的用户指纹，不会因用户指纹的状态发生变化，而出现难以识别的问题，识别的准确率较高，提升了用户体验。
需要说明的是，本说明书中的各个实施例均采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似的部分互相参见即可。
为了描述的方便，描述以上系统或装置时以功能分为各种模块或单元分别描述。当然，在实施本申请时可以把各单元的功能在同一个或多个软件和/或硬件中实现。
通过以上的实施方式的描述可知，本领域的技术人员可以清楚地了解到本申请可借助软件加必需的通用硬件平台的方式来实现。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例或者实施例的某些部分所述的方法。
最后，还需要说明的是，在本文中，诸如第一、第二、第三和第四等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从 而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。

Claims (17)

  1. 一种指纹识别方法，其特征在于，包括：
    获取用户输入的指纹；
    对所述指纹进行特征点提取，得到待识别的第一指纹特征；
    基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
    如果所述最优匹配度达到设定的阈值，则指纹识别成功。
  2. 根据权利要求1所述的方法，其特征在于，所述第一指纹特征、所述第二指纹特征分别包含相应个数的特征点，所述特征点包括终结点、分叉点、分歧点、孤立点、环点、短纹。
  3. 根据权利要求2所述的方法，其特征在于，所述基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度，包括：
    将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
    分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
    从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
  4. 根据权利要求3所述的方法，其特征在于，所述匹配结果包括n维0-1向量(a1,a2,......,an)，n为大于1的自然数，其中，
    ai＝1表示所述第一指纹特征中的第i个特征点，与所述第二指纹特征中的第i个特征点相匹配；
    ai＝0表示所述第一指纹特征中的第i个特征点，与所述第二指纹特征中的第i个特征点不匹配，i＝1,2,......,n。
  5. 根据权利要求4所述的方法，其特征在于，还包括：
    如果指纹识别成功，则依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
  6. 根据权利要求1所述的方法，其特征在于，还包括：
    预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
  7. 根据权利要求1所述的方法，其特征在于，还包括：
    如果所述最优匹配度未达到设定的阈值，则指纹识别失败。
  8. 一种指纹识别装置，其特征在于，包括：
    指纹获取模块，用于获取用户输入的指纹；
    特征提取模块，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征；
    匹配识别模块，用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
    第一结果判决模块，用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
  9. 根据权利要求8所述的装置，其特征在于，所述匹配识别模块包括：
    匹配单元，用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
    计算单元，用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
    选取单元，用于从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
  10. 根据权利要求9所述的装置，其特征在于，还包括：
    权值调整模块，用于在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
  11. 根据权利要求8所述的装置，其特征在于，还包括：
    预处理模块，用于预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
  12. 根据权利要求8所述的装置，其特征在于，还包括：
    第二结果判决模块，用于在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
  13. 一种电子设备，其特征在于，所述电子设备包括通信总线、收发装置、存储器以及处理器，其中：
    所述通信总线，用于实现所述收发装置、所述存储器以及所述处理器之间的连接通信；
    所述存储器中存储一组程序代码，且处理器调用存储器中存储的程序代码，用于执行以下操作：
    所述收发装置，用于获取用户输入的指纹；
    所述处理器，用于对所述指纹进行特征点提取，得到待识别的第一指纹特征；
    所述处理器，还用于基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度；其中，所述指纹特征数据库包含至少一个指纹模板，每个所述指纹模板包含第二指纹特征及P组特征点权值，每组特征点权值对应一用户指纹状态，且每组特征点权值用于计算其对应的用户指纹状态下用户指纹的匹配度数值，P为大于1的自然数；
    所述处理器，还用于在所述最优匹配度达到设定的阈值时，判决识别结果为指纹识别成功。
  14. 根据权利要求13所述的电子设备，其特征在于，所述处理器基于预设的指纹特征数据库对所述第一指纹特征进行匹配识别，并获取所述第一指纹特征的最优匹配度具体包括：
    所述处理器，还用于将所述第一指纹特征，与所述指纹特征数据库中每一指纹模板所包含的第二指纹特征进行匹配，得到匹配结果；
    所述处理器，还用于分别利用所述指纹模板的各组特征点权值对所述匹配结果进行加权计算，得到所述第一指纹特征在各种用户指纹状态下的匹配度；
    所述处理器，还用于从所述第一指纹特征对应于各指纹模板、各用户指纹状态的一系列匹配度中，选取出数值最大的匹配度，将其作为所述第一指纹特征的最优匹配度。
  15. 根据权利要求14所述的电子设备，其特征在于，所述处理器还用于执行如下步骤：
    在指纹识别成功时，依据计算出所述最优匹配度时所采用的匹配结果，对计算出所述最优匹配度时所采用的一组特征点权值进行调整。
  16. 根据权利要求13所述的电子设备，其特征在于，所述处理器还用于执行如下步骤：
    预先在预设的P种用户指纹状态下，对所述指纹模板进行特征点权值训练，得到与所述P种用户指纹状态一一对应的P组特征点权值。
  17. 根据权利要求13所述的电子设备，其特征在于，所述处理器还用于执行如下步骤：
    在所述最优匹配度未达到设定的阈值时，判决识别结果为指纹识别失败。
PCT/CN2015/082900 2015-06-17 2015-06-30 一种指纹识别方法、装置及电子设备 Ceased WO2016201731A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510337023.6 2015-06-17
CN201510337023.6A CN105654025B (zh) 2015-06-17 2015-06-17 一种指纹识别方法、装置及电子设备

Publications (1)

Publication Number Publication Date
WO2016201731A1 true WO2016201731A1 (zh) 2016-12-22

Family

ID=56481588

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/082900 Ceased WO2016201731A1 (zh) 2015-06-17 2015-06-30 一种指纹识别方法、装置及电子设备

Country Status (2)

Country Link
CN (1) CN105654025B (zh)
WO (1) WO2016201731A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106843987A (zh) * 2017-02-18 2017-06-13 珠海格力电器股份有限公司 基于指纹识别启动关联应用的方法及装置
CN114237383A (zh) * 2021-11-09 2022-03-25 浙江迈联医疗科技有限公司 基于前额单导联脑电信号的多状态识别方法
CN114915566A (zh) * 2021-01-28 2022-08-16 腾讯科技（深圳）有限公司 应用识别方法、装置、设备及计算机可读存储介质
CN116051865A (zh) * 2022-12-14 2023-05-02 杭州海康机器人股份有限公司 一种图像识别方法、模板库建立方法及装置

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106682618A (zh) * 2016-12-27 2017-05-17 努比亚技术有限公司 指纹识别方法及移动终端
CN106909900A (zh) * 2017-02-27 2017-06-30 努比亚技术有限公司 指纹识别方法及装置
WO2018213945A1 (zh) * 2017-05-20 2018-11-29 深圳信炜科技有限公司 图像传感器及电子装置
CN107330388B (zh) * 2017-06-21 2020-12-22 海信视像科技股份有限公司 指纹识别处理方法及终端
CN107480641B (zh) * 2017-08-16 2020-08-25 联想(北京)有限公司 一种指纹识别方法以及电子设备
CN109344594A (zh) * 2018-11-15 2019-02-15 Oppo（重庆）智能科技有限公司 一种基于指纹控制设备的方法及相关设备
KR20210129761A (ko) * 2020-04-20 2021-10-29 삼성디스플레이 주식회사 지문 감지 기능을 갖는 표시 장치 및 그것의 동작 방법
KR102822927B1 (ko) 2020-06-22 2025-06-23 삼성디스플레이 주식회사 지문 인증 장치, 이를 포함하는 표시 장치, 및 지문 인증 방법
CN115797978A (zh) * 2022-10-18 2023-03-14 北京合创芯彩科技有限公司 一种精准匹配指纹算法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101918981A (zh) * 2008-01-21 2010-12-15 日本电气株式会社 图案匹配系统、图案匹配方法以及图案匹配程序
CN104050406A (zh) * 2014-07-03 2014-09-17 南昌欧菲生物识别技术有限公司 利用指纹组合进行鉴权的方法及终端设备
CN104376245A (zh) * 2014-10-16 2015-02-25 上海嘉炽自动化系统工程有限公司 工业人机界面的操作权限确定方法及系统
CN104392162A (zh) * 2014-12-08 2015-03-04 北京大唐智能卡技术有限公司 一种身份验证方法及智能卡

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101499130B (zh) * 2008-01-30 2012-06-27 深圳市普罗巴克科技股份有限公司 一种指纹识别方法和指纹识别系统
CN101604384B (zh) * 2009-07-17 2012-06-27 周易 个性化指纹识别方法
CN103065134B (zh) * 2013-01-22 2016-01-13 江苏超创信息软件发展股份有限公司 一种具有提示信息的指纹识别装置和方法
CN103136517B (zh) * 2013-03-04 2015-11-11 杭州景联文科技有限公司 一种基于关键列选择的滚动指纹图像序列实时拼接方法
CN103324944B (zh) * 2013-06-26 2016-11-16 电子科技大学 一种基于svm和稀疏表示的假指纹检测方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101918981A (zh) * 2008-01-21 2010-12-15 日本电气株式会社 图案匹配系统、图案匹配方法以及图案匹配程序
CN104050406A (zh) * 2014-07-03 2014-09-17 南昌欧菲生物识别技术有限公司 利用指纹组合进行鉴权的方法及终端设备
CN104376245A (zh) * 2014-10-16 2015-02-25 上海嘉炽自动化系统工程有限公司 工业人机界面的操作权限确定方法及系统
CN104392162A (zh) * 2014-12-08 2015-03-04 北京大唐智能卡技术有限公司 一种身份验证方法及智能卡

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106843987A (zh) * 2017-02-18 2017-06-13 珠海格力电器股份有限公司 基于指纹识别启动关联应用的方法及装置
CN114915566A (zh) * 2021-01-28 2022-08-16 腾讯科技（深圳）有限公司 应用识别方法、装置、设备及计算机可读存储介质
CN114915566B (zh) * 2021-01-28 2024-05-17 腾讯科技（深圳）有限公司 应用识别方法、装置、设备及计算机可读存储介质
CN114237383A (zh) * 2021-11-09 2022-03-25 浙江迈联医疗科技有限公司 基于前额单导联脑电信号的多状态识别方法
CN114237383B (zh) * 2021-11-09 2024-03-12 浙江迈联医疗科技有限公司 基于前额单导联脑电信号的多状态识别方法
CN116051865A (zh) * 2022-12-14 2023-05-02 杭州海康机器人股份有限公司 一种图像识别方法、模板库建立方法及装置

Also Published As

Publication number Publication date
CN105654025A (zh) 2016-06-08
CN105654025B (zh) 2019-01-25

Similar Documents

Publication Publication Date Title
WO2016201731A1 (zh) 一种指纹识别方法、装置及电子设备
US10198616B2 (en) Method for fingerprint unlocking and terminal
US20200134145A1 (en) Apparatus and methods for biometric verification
WO2019119505A1 (zh) 人脸识别的方法和装置、计算机装置及存储介质
US11893831B2 (en) Identity information processing method and device based on fundus image
CN104820983B (zh) 一种图像匹配方法
US8995729B2 (en) Accelerated comparison using scores from coarse and fine matching processes
WO2017162076A1 (zh) 一种人脸识别方法及系统
US11062120B2 (en) High speed reference point independent database filtering for fingerprint identification
CN109508658B (zh) 一种多生物特征联合识别方法及装置
EP2289020A2 (en) Fingerprint representation using gradient histograms
KR102415504B1 (ko) 사용자 인증을 위한 등록 데이터베이스의 갱신 방법 및 장치
US20100033304A1 (en) Identification Method Control Device and Identification Method Control Method
CN106326327A (zh) 用于更新用户认证数据的方法和设备
TWI437501B (zh) 基於生物特徵之身分驗證裝置及其方法
CN114464193A (zh) 声纹聚类方法、装置、存储介质及电子装置
US20190244006A1 (en) Skin information processing method, skin information processing device, and non-transitory computer-readable medium
CN110659541A (zh) 图像识别方法、装置及存储介质
CN106650554A (zh) 静态手势识别方法
CN112861742B (zh) 人脸识别方法、装置、电子设备及存储介质
US11893825B2 (en) Method for determining a match between a candidate fingerprint and a reference fingerprint
JP2009503721A (ja) 指紋のアライニング方法および装置
US20200387691A1 (en) A quick match algorithm for biometric data
CN102385693B (zh) 手指静脉辨识系统与方法
Liu et al. Finger-vein recognition with modified binary tree model

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: 15895304

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: 15895304

Country of ref document: EP

Kind code of ref document: A1