WO2011095071A1 - 一种用户认证的方法及装置 - Google Patents

一种用户认证的方法及装置 Download PDF

Info

Publication number
WO2011095071A1
WO2011095071A1 PCT/CN2011/070222 CN2011070222W WO2011095071A1 WO 2011095071 A1 WO2011095071 A1 WO 2011095071A1 CN 2011070222 W CN2011070222 W CN 2011070222W WO 2011095071 A1 WO2011095071 A1 WO 2011095071A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
photos
authentication
predetermined
saved
Prior art date
Application number
PCT/CN2011/070222
Other languages
English (en)
French (fr)
Inventor
胡子敏
吴昊
Original Assignee
腾讯科技(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 腾讯科技(深圳)有限公司 filed Critical 腾讯科技(深圳)有限公司
Priority to SG2012054821A priority Critical patent/SG182702A1/en
Publication of WO2011095071A1 publication Critical patent/WO2011095071A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172Classification, e.g. identification

Definitions

  • the invention relates to a method and a device for user authentication, belonging to the technical field of network communication. Background of the invention
  • the present invention provides a method and device for user authentication, which solves the problem that the manual discriminating method is slow in the prior art and the user waits for a long authentication time. Therefore, the present invention adopts the following technical solution:
  • the embodiment of the invention provides a method for user authentication, including:
  • the saved photos of the same user are identified according to a predetermined template, and if the predetermined requirements are met, a plurality of saved photos of the user are randomly selected and sent to a designated plurality of users for judgment or scoring;
  • the predetermined requirement is not met, or more than a predetermined number of designated users, the multiple photos of the user who are randomly selected and saved are not the same user's photo, or exceed the predetermined number of designations. If the user scores a plurality of photos of the user that are randomly selected and saved according to a specific standard, the score is less than a specific value, and the authenticated user is illegal.
  • the embodiment of the invention further provides a device for user authentication, including:
  • a predetermined template identification unit configured to identify the saved photo of the same user according to a predetermined template, and if the predetermined requirement is met, execute a random sending unit, and if the predetermined requirement is not met, the authentication user is invalid;
  • a random sending unit configured to randomly extract and save a plurality of photos of the user to be sent to a plurality of specified users for judgment or scoring;
  • Receiving an authentication unit configured to receive a determination result or a score reply of the specified user sent by the random sending unit, if more than a predetermined number of specified users determine that the plurality of photos of the user that are randomly selected and saved are photos of the same user, or If the predetermined number of designated users exceeds a certain value for a plurality of photos of the user that are randomly selected and saved according to a specific standard, the authenticated user is legal, and if the specified number of specified users exceeds the predetermined number of specified users, the user is randomly selected and saved.
  • the plurality of photos are judged not to be photos of the same user, or more than a predetermined number of designated users have scored less than a specific value for a plurality of randomly selected photos of the user, according to a specific criterion, and the authenticated user is illegal.
  • the technical solution provided by the invention improves the user's working process by the preliminary screening of the predetermined template and the filtering of the designated users by the designated user, and improves the speed of the user authentication and meets the needs of the user.
  • FIG. 1 is a schematic flowchart of a method for user authentication according to an embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of an optional template in a method for user authentication according to an embodiment of the present invention
  • FIG. 3 is a schematic flowchart diagram of a method for user authentication according to an example of the embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of an apparatus for user authentication according to an embodiment of the present invention. Mode for carrying out the invention
  • the present application is mainly applied to the instant messaging service of the present invention as a preferred embodiment.
  • the actual application can also be applied to other Internet service systems such as a network mail service system and a network collaborative work service system.
  • a method for user authentication provided by a specific embodiment of the present invention, as shown in FIG. 1, may specifically include the following steps:
  • Step 11 The saved photos of the same user are identified according to a predetermined template. If the predetermined requirements are met, step 12 is performed; if the predetermined requirements are not met, step 16 is performed; and step 12, the saved users are randomly selected. After the photo is sent to the specified multiple users, perform step 13 or step 14;
  • Step 13 it is determined that more than a predetermined number of designated users determine whether the multiple photos of the user are randomly selected to be the same user's photo, and if so, step 15 is performed, if no, step 16 is performed;
  • Step 14 Determine whether a score exceeds a certain value by a predetermined number of specified users for a plurality of photos of the user that are randomly selected and saved according to a specific criterion, and if yes, execute step 15, and if not, execute step 16;
  • Step 15. The authenticated user is legal; Step 16. The authenticated user is not legal.
  • the predetermined template is a feature vector that uses a rectangle as a face detection, and distinguishes a feature area in the photo by a color value in the rectangle.
  • the predetermined requirement means that the feature area of the photo in the predetermined template is set according to the color value of the face. set.
  • the eigenvalues of the feature vector can be defined as a white rectangular pigment and a black rectangle pigment. The eye part is darker than the cheek part.
  • the sides of the nose are darker than the nose, and the mouth is darker than the surrounding area.
  • the user may continuously capture a plurality of photos of the user and save them, or receive a plurality of photos sent by the user and save them.
  • the specific standard, the predetermined quantity and the specific value are determined according to the security level of the certification requirement, and the corresponding specific standard may specifically include: Cannot be a cartoon character avatar, a star character avatar or a child character avatar, cannot be a multi-person photo , the face can not be occluded or incomplete, can not be a side avatar.
  • the specific standard may only be a cartoon character avatar, a star avatar or a child avatar; the predetermined number may be set to 3 or more, and the specific value may be set only to 60 points or more (total score 100 points); If the certification requirements are highly secure, then the specific criteria need to be set to not be a cartoon character avatar, star avatar or child avatar, and not a multi-person photo. The face must not be occluded or incomplete, not a side avatar, etc.; the predetermined number needs to be set to 10 or above, and the specific value needs to be set to 70 points or more (total score of 100 points).
  • the specified multiple users can be authenticated devices or staff members based on the user's identity information. For those who are high-level or responsible, for example, if the user's level is primary but can report some offending users or actively help other users, then the user can be specified, or the user's level is advanced and always operated by chapter Then, the user can also be specified; the specified multiple users can also be determined according to the relationship with the authenticated user. For example, a total number of users less than the specified number of users can be designated as the authenticated user's friends, for example, a total designation. 6 users, then the authenticated user's friends can't be more than 3, only 2 or 1 person, increase the authenticated user's friend for authentication, can clearly understand whether the user's photo is his or her own, and less than half guarantee the certification. Fairness.
  • Step 31 continuously capture multiple photos of the same user through video and save them, or receive multiple photos sent by the user and save them;
  • Step 32 the saved photo of the same user is determined according to a predetermined template as a person's face, and if so, step 33 is performed, and if not, step 36 is performed;
  • Step 33 Randomly extract a plurality of photos of the user saved in step 31 and send them to a specified plurality of users for judgment or scoring;
  • Step 34 Receive a decision result or a score reply of the specified multiple users, and determine whether a predetermined number of specified users exceeds a plurality of photos of the user that are randomly selected and saved as a photo of the same user, or determine that the predetermined number is exceeded.
  • the designated user according to a specific standard for randomly extracting the saved scores of the plurality of photos of the user is greater than or equal to a specific value, and if so, executing step 35; if not, executing step 36;
  • Step 35 The authenticated user is legal
  • Step 36 The authenticated user is invalid.
  • the plurality of photos of the user that are randomly selected and saved may be sent to the staff for another review. If the audit result is still unqualified, the user is notified that the authentication is illegal, and if the result of the audit is If it is legal, the authenticated user is legal.
  • the technical solution provided by the embodiment can compress the workload of the auditor, and the preliminary screening can be performed by judging whether the saved photo of the same user is in accordance with a predetermined template according to a predetermined template, and the user who is obviously unqualified by the face recognition technology can be mainly used.
  • the child character avatar and the like notify the administrator; Finally, the staff member verifies the photos of some of the reported users who have been disputed, and then authenticates the user.
  • the embodiment of the present invention further provides a technical solution for a device for user authentication. As shown in FIG. 4, the method includes:
  • the predetermined template identification unit 41 is configured to identify the saved photos of the same user according to a predetermined template, and if the predetermined requirements are met, execute the random sending unit 42. If the predetermined requirements are not met, the authentication user is illegal;
  • the random sending unit 42 is configured to randomly extract and save the saved plurality of photos of the user to the designated plurality of users for judgment or scoring;
  • the receiving authentication unit 43 is configured to receive a determination result or a score reply of the specified user sent by the random sending unit 42, and if the predetermined number of specified users exceeds the multiple photos of the user that are randomly selected and saved, the photo is the same user, or More than a predetermined number of designated users scored a plurality of photos of the user's multiple photos that were randomly selected and saved according to a specific standard. For a specific value, the authenticated user is legal. If more than a predetermined number of designated users decide that multiple photos of the user that are randomly selected and saved are not photos of the same user, or more than a predetermined number of designated users are randomly selected according to a specific standard. If the scores of the plurality of photos of the user are less than a specific value, the authenticated user is illegal.
  • the above device may further include:
  • a storage unit that saves multiple photos of a user who are continuously shot by video, or saves multiple photos of the received user.
  • the predetermined template identification unit 41 is configured to identify a plurality of photos of the same user saved in the storage unit according to a predetermined template, and if the predetermined requirement is met, execute the random sending unit 42 to authenticate the user if the predetermined requirement is not met. illegal.
  • the predetermined template in the predetermined template identification unit 41 is a feature vector that uses a rectangle as a face detection, and distinguishes a feature region in the photo by a color value in the rectangle, where the predetermined requirement refers to a predetermined template.
  • the feature area of the photo is set according to the color value of the face.
  • the eigenvalues of the feature vector can be defined as a white rectangular pigment and a black rectangle pigment.
  • the eye part is darker than the cheek part.
  • the sides of the nose are darker than the nose, and the mouth is darker than the surrounding area.
  • the black portion of the rectangle in Figure 2a is placed in the eye or mouth portion of the person in the photo, and the white portion of the rectangle in Figure 2b is placed in the nose of the person in the photo. It is also possible to use other templates for recognizing a person's face, as long as it can identify the characteristic parts of the person: Yes.
  • the specific standard, the predetermined quantity, and the specific value in the receiving authentication unit 43 are all determined according to the security level of the authentication requirement, and the specific standard specifically includes: cannot be a cartoon character avatar, a star avatar, or a child avatar, cannot It is a multi-person photo, the face can not be occluded or incomplete, can not be a side avatar.
  • the specific standard may only be a cartoon character avatar, a star avatar or a child avatar; the predetermined number may be 3 or more, and the specific value may be only 60 or more.
  • the specific value needs to be set to 70 points or more (total score of 100 points).
  • the specified multiple users may be authenticated by the authentication device or the staff according to the user's identity information, for example, the user's level is primary but can report some offending users or actively help other users. Then the user can be specified, or the user's level is advanced and always operated by chapter, then the user can also be specified; the specified multiple users can also be determined according to the relationship with the authenticated user, for example, the total number of people Less than half of the specified users are designated as friends of the authenticated user. For example, if a total of 6 users are specified, the authenticated user's friends cannot exceed 3, and only 2 or 1 people can be added. A clear understanding of whether the user's photo is his or her own, and less than half guarantees the fairness of the certification.
  • the authenticated user After the authenticated user is illegal, it can also include:
  • the notification unit is configured to notify the staff to perform another review on the plurality of photos of the user that are randomly selected, and if the audit result is still unsatisfactory, notify the user that the authentication is illegal, and if the audit result is legal, the user authentication is legal.
  • the predetermined template identification unit 41 performs the initial screening of the saved photos of the same user according to a predetermined template, and the main screening can be performed by the face recognition.
  • the technology notifies the staff after the user authentication user who is obviously unqualified is illegal, and secondly, the user is scored by the user who randomly selects a plurality of photos of the user, and judges whether the judgment is determined in the discriminating process by judging that the predetermined number of users exceeds the predetermined number. Screening out some star characters, cards, and photos of the same user or whether the scores scored during the user scoring process are greater than or equal to a specific value. The avatar or the avatar of the child is notified to the administrator.
  • the staff member verifies the photos of some of the reported users who have been reported, and then authenticates the user, thus reducing the workload of the staff, through preliminary screening,
  • the process of discriminating or scoring the user saves the auditing cost and provides convenience for the authenticated user. It also ensures the security accuracy of the user authentication, enables the legal authenticated user to pass the authentication in the shortest time, and improves the speed of the user authentication. User needs.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Collating Specific Patterns (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Description

一种用户认证的方法及装置
本申请要求 2010年 2月 2 日在中国知识产权局递交的中国专利申 请 No. 2010101104226的优先权, 其全部内容通过引用合并于此。 技术领域
本发明涉及一种用户认证的方法及装置, 属于网络通讯技术领域。 发明背景
目前用户认证的方法都是人工进行审核, 审核人员工作量大, 有些 合法的用户要好多天之后才能通过认证, 给用户带来了不便。 发明内容
本发明提供了一种用户认证的方法及装置, 以解决现有技术中存在 的人工判别方式较慢、 用户等待认证时间长的问题, 为此本发明采用如 下的技术方案:
本发明实施例提供了一种用户认证的方法, 包括:
将保存的同一用户的照片按照预定的模板进行识别, 若符合预定要 求, 则随机抽取保存的所述用户的多张照片发送给指定的多个用户进行 判决或打分;
若超过预定数量的指定用户对随机抽取保存的所述用户的多张照 片判决是同一用户的照片, 或超过预定数量的指定用户按照特定的标准 对随机抽取保存的所述用户的多张照片打出的分数大于等于特定值, 则 认证用户合法;
若不符合预定要求, 或超过预定数量的指定用户对随机抽取保存的 所述用户的多张照片判决不是同一用户的照片, 或超过预定数量的指定 用户按照特定的标准对随机抽取保存的所述用户的多张照片打出的分 数小于特定值, 则认证用户不合法。
本发明实施例还提供了一种用户认证的装置, 包括:
预定模板识别单元, 用于将保存的同一用户的照片按照预定的模板 进行识别, 若符合预定要求, 则执行随机发送单元, 若不符合预定要求, 则认证用户不合法;
随机发送单元, 用于随机抽取保存的所述用户的多张照片发送给指 定的多个用户进行判决或打分;
接收认证单元, 用于接收所述随机发送单元发送的指定用户的判决 结果或分数回复, 若超过预定数量的指定用户对随机抽取保存的所述用 户的多张照片判决是同一用户的照片, 或超过预定数量的指定用户按照 特定的标准对随机抽取保存的所述用户的多张照片打出的分数大于等 于特定值, 则认证用户合法, 若超过预定数量的指定用户对随机抽取保 存的所述用户的多张照片判决不是同一用户的照片, 或超过预定数量的 指定用户按照特定的标准对随机抽取保存的所述用户的多张照片打出 的分数小于特定值, 则认证用户不合法。
本发明提供的技术方案通过预定模板的初步筛选和指定用户对初步 筛选的合格用户的过滤, 筒化了工作人员的工作过程, 提高了用户认证 的速度, 满足了用户的需求。 附图简要说明
为了更清楚地说明本发明实施例的技术方案, 下面将对实施例描述 中所需要使用的附图作筒单地介绍, 显而易见地, 下面描述中的附图仅 仅是本发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创 造性劳动性的前提下, 还可以根据这些附图获得其他的附图。 图 1是本发明实施例所述的一种用户认证的方法的流程示意图; 图 2a和图 2b是本发明实施例所述的一种用户认证的方法中可选的预 定模板的结构示意图;
图 3是本发明实施例所述的举例具体说明的一种用户认证的方法的 流程示意图;
图 4是本发明实施例所述的一种用户认证的装置的结构示意图。 实施本发明的方式
下面结合说明书附图来说明本发明的具体实施方式。 本说明书主要 以本发明在即时通信服务中的应用作为最佳实施例, 当然, 实际应用中 也可以用于网络邮件服务系统、 网络协同工作服务系统等其它互联网服 务系统。
本发明的具体实施方式提供的一种用户认证的方法, 如图 1所示, 具 体可以包括如下的步骤:
步骤 11 , 将保存的同一用户的照片按照预定的模板进行识别, 若符 合预定要求, 则执行步骤 12; 若不符合预定要求, 则执行步骤 16; 步骤 12, 随机抽取保存的所述用户的多张照片发送给指定的多个 用户后, 执行步骤 13或步骤 14;
步骤 13、 判断超过预定数量的指定用户对随机抽取保存的所述用户 的多张照片判决是否为同一用户的照片, 若是, 则执行步骤 15, 若否, 则执行步骤 16;
步骤 14、 判断超过预定数量的指定用户按照特定的标准对随机抽取 保存的所述用户的多张照片打出的分数是否大于等于特定值, 若是, 则 执行步骤 15, 若否, 则执行步骤 16;
步骤 15、 认证用户合法; 步骤 16、 认证用户不合法。
上述方法中, 认证用户不合法后还可以将该随机抽取保存的所述用 户的多张照片发送给工作人员再进行一次审核, 若审核结果仍不合格, 则通知用户认证不合法, 若审核结果合法, 则认证用户合法。 预定的模 板是将矩形作为人脸检测的特征向量, 并通过矩形中的色值区别出照片 中的特征区域, 所述预定要求, 是指预定模板中照片的特征区域根据人 脸的色值设定。 例如: 可以将特征向量的特征值定义为白色矩形色素和 减去黑色矩形色素和, 眼睛部分要比脸颊部分颜色更深, 鼻梁两侧要比 鼻梁颜色更深, 嘴巴要比其周围部位颜色更深, 那么, 就用图 2 ( a ) 中 的矩形的黑色部分放置于照片中人物的眼睛或嘴巴部位, 用图 2 ( b ) 中 的矩形的白色部分放置于照片中人物的鼻梁部位。 用其它识别人脸的模 板也可以, 只要能识别出是人物的特征部位就可以。
上述方法中, 在步骤 11之前, 还可以通过视频连续拍摄用户的多张 照片并保存, 或接收用户发送的多张照片后保存。
进一步, 特定的标准、 预定数量和特定值都是根据认证要求的安全 程度确定的, 相应的特定的标准可以具体包括: 不能是卡通人物头像、 明星人物头像或小孩人物头像, 不能是多人合照, 脸部不能有遮挡或不 完整, 不能是侧面头像等。 例如: 若认证要求的安全程度较低, 那么特 定的标准可以仅为不能是卡通人物头像、 明星人物头像或小孩人物头 像; 预定数量可以设定为 3人或以上, 特定值可以仅设定为 60分或以上 (总分 100分); 若认证要求的安全程度很高, 那么特定的标准就需要设 定为不能是卡通人物头像、 明星人物头像或小孩人物头像, 且不能是多 人合照, 脸部不能有遮挡或不完整, 不能是侧面头像等; 预定数量就需 要设定为 10或以上, 特定值就需要设定为 70分或以上(总分 100分)。
指定的多个用户可以是认证装置或工作人员根据用户的身份信息从 级别高或责任心强的用户中挑取的, 例如, 用户的级别为初级但是能够 举报一些违规用户或积极帮助其他用户, 那么该用户可以被指定, 或者 用户的级别为高级且一直按章操作, 那么该用户也可以被指定; 指定的 多个用户也可以是根据与认证用户的关系确定的, 例如, 可以将总人数 少于指定的多个用户一半指定为认证用户的好友, 比如总共指定 6个用 户, 那么认证用户的好友不能超过 3人, 仅能是 2人或 1人, 增加认证用 户的好友进行认证, 能够清楚的了解用户的照片是否是其本人, 而少于 半数保证了认证的公平性。
综合上述方法, 如图 3所示, 举例具体说明用户认证的全过程: 步骤 31、 通过视频连续拍摄同一用户的多张照片并保存, 或接收用 户发送的多张照片后保存;
步骤 32、 将保存的同一用户的照片按照预定的模板判断是否为人物 脸部, 若是, 则执行步骤 33 , 若否, 执行步骤 36;
步骤 33、 随机抽取步骤 31中保存的所述用户的多张照片发送给指定 的多个用户进行判决或打分;
步骤 34、 接收所述指定的多个用户的判决结果或分数回复, 判断超 过预定数量的指定用户对随机抽取保存的所述用户的多张照片判决是 否为同一用户的照片, 或判断超过预定数量的指定用户按照特定的标准 对随机抽取保存的所述用户的多张照片打出的分数是否大于等于特定 值, 若是, 则执行步骤 35; 若否, 则执行步骤 36;
步骤 35、 认证用户合法;
步骤 36、 认证用户不合法。
上述方法中, 认证用户不合法后还可以将该随机抽取保存的所述用 户的多张照片发送给工作人员再进行一次审核, 若审核结果仍不合格, 则通知用户认证不合法, 若审核结果合法, 则认证用户合法。 本实施方式提供的技术方案可以筒化审核人员的工作量, 通过将保 存的同一用户的照片按照预定的模板判断是否符合预定要求进行初步 筛选, 主要可以通过人脸识别技术将明显不合格的用户认证用户不合法 后通知工作人员; 然后, 通过指定用户对随机抽取保存的所述用户的多 张照片的用户判别过程, 或者指定用户对随机抽取保存的所述用户的多 张照片的用户打分过程, 并且通过判断超过预定数量的指定用户在判别 过程中判决是否为同一用户的照片或在用户打分过程中打出的分数是 否大于等于特定值的判断过程, 筛选出一些明星人物头像、 卡通人物头 像或小孩人物头像等通知管理员; 最后, 工作人员对上报的一些存在争 议的用户的照片进行审核后, 实现对用户的认证。 这样筒化了工作人员 的工作量, 通过初步筛选、 用户判别或打分的过程, 节约了审核成本, 为认证用户提供了便利; 并且保证了用户认证的安全准确性, 使合法认 证用户能够在最短时间内通过认证, 提高了用户认证的速度, 满足了用 户的需求。
本发明的具体实施方式还提供了一种用户认证的装置的技术方案, 如图 4所示, 包括:
预定模板识别单元 41 , 用于将保存的同一用户的照片按照预定的模 板进行识别, 若符合预定要求, 则执行随机发送单元 42, 若不符合预定 要求, 则认证用户不合法;
随机发送单元 42, 用于随机抽取保存的所述用户的多张照片发送给 指定的多个用户进行判决或打分;
接收认证单元 43 , 用于接收随机发送单元 42发送的指定用户的判决 结果或分数回复, 若超过预定数量的指定用户对随机抽取保存的所述用 户的多张照片判决是同一用户的照片, 或超过预定数量的指定用户按照 特定的标准对随机抽取保存的所述用户的多张照片打出的分数大于等 于特定值, 则认证用户合法, 若超过预定数量的指定用户对随机抽取保 存的所述用户的多张照片判决不是同一用户的照片, 或超过预定数量的 指定用户按照特定的标准对随机抽取保存的所述用户的多张照片打出 的分数小于特定值, 则认证用户不合法。
上述装置中, 还可以包括:
存储单元, 用于将通过视频连续拍摄的用户多张照片进行保存, 或 将接收到的用户的多张照片进行保存。
所述预定模板识别单元 41 , 用于将存储单元中保存的同一用户的多 张照片按照预定的模板进行识别, 若符合预定要求, 则执行随机发送单 元 42, 若不符合预定要求, 则认证用户不合法。
进一步, 预定模板识别单元 41中所述预定的模板, 是将矩形作为人 脸检测的特征向量, 并通过矩形中的色值区别出照片中的特征区域, 所 述预定要求, 是指预定模板中照片的特征区域根据人脸的色值设定。 例 如: 可以将特征向量的特征值定义为白色矩形色素和减去黑色矩形色素 和, 眼睛部分要比脸颊部分颜色更深, 鼻梁两侧要比鼻梁颜色更深, 嘴 巴要比其周围部位颜色更深, 那么, 就用图 2a中的矩形的黑色部分放置 于照片中人物的眼睛或嘴巴部位, 用图 2b中的矩形的白色部分放置于照 片中人物的鼻梁部位。 用其它识别人脸的模板也可以, 只要能识别出是 人物的特征部位: 可以。
接收认证单元 43中所述特定的标准、预定数量和特定值都是根据认 证要求的安全程度确定的, 所述特定的标准具体包括: 不能是卡通人物 头像、 明星人物头像或小孩人物头像, 不能是多人合照, 脸部不能有遮 挡或不完整, 不能是侧面头像。 例如: 若认证要求的安全程度较低, 那 么特定的标准可以仅为不能是卡通人物头像、 明星人物头像或小孩人物 头像; 预定数量可以为 3人或以上, 特定值可以仅为 60分或以上(总分 100分); 若认证要求的安全程度很高, 那么特定的标准就需要设定为不 能是卡通人物头像、明星人物头像或小孩人物头像,且不能是多人合照, 脸部不能有遮挡或不完整, 不能是侧面头像等; 预定数量就需要设定为
10或以上, 特定值就需要设定为 70分或以上(总分 100分)。
指定的多个用户可以是认证装置或工作人员根据用户的身份信息 从级别高或责任心强的用户中挑取的, 例如, 用户的级别为初级但是能 够举报一些违规用户或积极帮助其他用户, 那么该用户可以被指定, 或 者用户的级别为高级且一直按章操作, 那么该用户也可以被指定; 指定 的多个用户也可以是根据与认证用户的关系确定的, 例如, 可以是总人 数少于指定的多个用户一半指定为认证用户的好友, 比如总共指定 6个 用户, 那么认证用户的好友不能超过 3人, 仅能是 2人或 1人, 增加认证 用户的好友进行认证, 能够清楚的了解用户的照片是否是其本人, 而少 于半数保证了认证的公平性。
认证用户不合法后还可以包括:
通知单元, 用于通知工作人员对该随机抽取保存的所述用户的多张 照片再进行一次审核, 若审核结果仍不合格, 则通知用户认证不合法, 若审核结果合法, 则用户认证合法。
本发明实施例提供了一种用户认证的装置的技术方案, 首先, 通过 预定模板识别单元 41将保存的同一用户的照片按照预定的模板进行识 别完成对用户的初步筛选, 主要可以通过人脸识别技术将明显不合格的 用户认证用户不合法后通知工作人员, 其次, 通过指定用户对随机抽取 用户的多张照片的用户打分过程, 并且通过判断超过预定数量的指定用 户在判别过程中判决是否为同一用户的照片或在用户打分过程中打出 的分数是否大于等于特定值的判断过程, 筛选出一些明星人物头像、 卡 通人物头像或小孩人物头像等通知管理员, 最后, 工作人员对上报的一 些存在争议的用户的照片进行审核后, 实现对用户的认证, 这样筒化了 工作人员的工作量, 通过初步筛选、 用户判别或打分的过程, 节约了审 核成本, 为认证用户提供了便利; 并且保证了用户认证的安全准确性, 使合法认证用户能够在最短时间内通过认证, 提高了用户认证的速度, 满足了用户的需求。
上述装置中包含的各单元的处理功能的具体实现方式在之前的方 法实施例中已经描述, 在此不再重复描述。
以上所述, 仅为本发明较佳的具体实施方式, 但本发明的保护范围 并不局限于此, 任何熟悉本技术领域的技术人员在本发明揭露的技术范 围内, 可轻易想到的变化或替换, 都应涵盖在本发明的保护范围之内。 因此, 本发明的保护范围应该以权利要求书的保护范围为准。

Claims

权利要求书
1、 一种用户认证的方法, 其特征在于, 包括:
将保存的同一用户的照片按照预定的模板进行识别, 若符合预定要 求, 则随机抽取保存的所述用户的多张照片发送给指定的多个用户进行 判决或打分;
若超过预定数量的指定用户对随机抽取保存的所述用户的多张照 片判决是同一用户的照片, 或超过预定数量的指定用户按照特定的标准 对随机抽取保存的所述用户的多张照片打出的分数大于等于特定值, 则 认证用户合法;
若不符合预定要求, 或超过预定数量的指定用户对随机抽取保存的 所述用户的多张照片判决不是同一用户的照片, 或超过预定数量的指定 用户按照特定的标准对随机抽取保存的所述用户的多张照片打出的分 数小于特定值, 则认证用户不合法。
2、 根据权利要求 1所述的用户认证的方法, 其特征在于, 在将保存 的同一用户的照片按照预定的模板进行识别之前, 还包括:
通过视频连续拍摄用户的多张照片并保存, 或接收用户发送的多张 照片后保存。
3、 根据权利要求 1所述的用户认证的方法, 其特征在于, 所述预定 的模板, 是将矩形作为人脸检测的特征向量, 并通过矩形中的色值区别 出照片中的特征区域, 所述预定要求, 是指预定模板中照片的特征区域 根据人脸的色值设定。
4、 根据权利要求 1所述的用户认证的方法, 其特征在于, 所述特定 的标准、 预定数量和特定值都是根据认证要求的安全程度确定的, 所述 特定的标准具体包括: 不能是卡通人物头像、 明星人物头像或小孩人物 头像, 不能是多人合照, 脸部不能有遮挡或不完整, 不能是侧面头像。
5、 根据权利要求 1所述的用户认证的方法, 其特征在于, 所述指定 的多个用户是根据用户的身份信息的级别或与认证用户的关系确定的。
6、根据权利要求 1-5中任一项所述的用户认证的方法, 其特征在于, 在认证用户不合法后进一步包括将该随机抽取保存的所述用户的多张 照片发送给工作人员进行审核, 若审核结果不合格, 则通知用户认证不 合法, 若审核结果合法, 则认证用户合法。
7、 一种用户认证的装置, 其特征在于, 包括:
预定模板识别单元, 用于将保存的同一用户的照片按照预定的模板 进行识别, 若符合预定要求, 则执行随机发送单元, 若不符合预定要求, 则认证用户不合法;
随机发送单元, 用于随机抽取保存的所述用户的多张照片发送给指 定的多个用户进行判决或打分;
接收认证单元, 用于接收所述随机发送单元发送的指定用户的判决 结果或分数回复, 若超过预定数量的指定用户对随机抽取保存的所述用 户的多张照片判决是同一用户的照片, 或超过预定数量的指定用户按照 特定的标准对随机抽取保存的所述用户的多张照片打出的分数大于等 于特定值, 则认证用户合法, 若超过预定数量的指定用户对随机抽取保 存的所述用户的多张照片判决不是同一用户的照片, 或超过预定数量的 指定用户按照特定的标准对随机抽取保存的所述用户的多张照片打出 的分数小于特定值, 则认证用户不合法。
8、 根据权利要求 7所述的用户认证的装置, 其特征在于, 在预定模 板识别单元将保存的同一用户的照片按照预定的模板进行识别之前, 还 包括:
存储单元, 用于将通过视频连续拍摄的用户多张照片进行保存, 或 将接收到的用户的多张照片进行保存。
9、 根据权利要求 7所述的用户认证的装置, 其特征在于, 所述预定 模板识别单元中所述预定的模板, 是将矩形作为人脸检测的特征向量, 并通过矩形中的色值区别出照片中的特征区域, 所述预定要求, 是指预 定模板中照片的特征区域根据人脸的色值设定。
10、 根据权利要求 7所述的用户认证的装置, 其特征在于, 所述接 收认证单元中所述特定的标准、 预定数量和特定值都是根据认证要求的 安全程度确定的, 所述特定的标准具体包括: 不能是卡通人物头像、 明 星人物头像或小孩人物头像, 不能是多人合照, 脸部不能有遮挡或不完 整, 不能是侧面头像。
11、 根据权利要求 7所述的用户认证的装置, 其特征在于, 所述随 机发送单元中所述指定的多个用户是根据用户的身份信息的级别或与 认证用户的关系确定的。
12、根据权利要求 7-11中任一项所述的用户认证的装置, 其特征在 于, 进一步包括通知单元, 用于在认证用户不合法后通知工作人员对该 随机抽取保存的所述用户的多张照片进行审核, 若审核结果不合格, 则 通知用户认证不合法, 若审核结果合法, 则用户认证合法。
PCT/CN2011/070222 2010-02-02 2011-01-13 一种用户认证的方法及装置 WO2011095071A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
SG2012054821A SG182702A1 (en) 2010-02-02 2011-01-13 Method and apparatus for user authentication

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201010110422.6A CN102142087B (zh) 2010-02-02 2010-02-02 一种用户认证的方法及装置
CN201010110422.6 2010-02-02

Publications (1)

Publication Number Publication Date
WO2011095071A1 true WO2011095071A1 (zh) 2011-08-11

Family

ID=44354950

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/070222 WO2011095071A1 (zh) 2010-02-02 2011-01-13 一种用户认证的方法及装置

Country Status (4)

Country Link
CN (1) CN102142087B (zh)
MY (1) MY164619A (zh)
SG (1) SG182702A1 (zh)
WO (1) WO2011095071A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112333239A (zh) * 2020-10-10 2021-02-05 百度(中国)有限公司 业务审核通知方法及网关、电子设备、可读介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104751033B (zh) * 2015-03-03 2017-11-24 上海拍拍贷金融信息服务有限公司 一种基于音视频文件的用户行为认证方法和装置
CN105608354A (zh) * 2015-07-01 2016-05-25 南京酷派软件技术有限公司 鉴权方法、鉴权系统、终端和服务器
CN111770279B (zh) * 2020-08-03 2022-04-08 维沃移动通信有限公司 一种拍摄方法及电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040181552A1 (en) * 2003-03-14 2004-09-16 Milne Donald A. Method and apparatus for facial identification enhancement
CN1687961A (zh) * 2005-04-19 2005-10-26 浙江大学 一种计算机辅助的国画真伪鉴别方法
US20090060288A1 (en) * 2005-09-28 2009-03-05 Charles A Myers Image Classification And Information Retrieval Over Wireless Digital Networks And The Internet
CN101482919A (zh) * 2008-01-08 2009-07-15 欧姆龙株式会社 人脸核对装置

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008262549A (ja) * 2007-03-16 2008-10-30 Ricoh Co Ltd 認証方法及び認証装置
CN101114909B (zh) * 2007-08-17 2011-02-16 上海博康智能信息技术有限公司 全自动视频身份认证系统及方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040181552A1 (en) * 2003-03-14 2004-09-16 Milne Donald A. Method and apparatus for facial identification enhancement
CN1687961A (zh) * 2005-04-19 2005-10-26 浙江大学 一种计算机辅助的国画真伪鉴别方法
US20090060288A1 (en) * 2005-09-28 2009-03-05 Charles A Myers Image Classification And Information Retrieval Over Wireless Digital Networks And The Internet
CN101482919A (zh) * 2008-01-08 2009-07-15 欧姆龙株式会社 人脸核对装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112333239A (zh) * 2020-10-10 2021-02-05 百度(中国)有限公司 业务审核通知方法及网关、电子设备、可读介质
CN112333239B (zh) * 2020-10-10 2023-07-18 百度(中国)有限公司 业务审核通知方法及网关、电子设备、可读介质

Also Published As

Publication number Publication date
SG182702A1 (en) 2012-08-30
CN102142087B (zh) 2015-09-09
MY164619A (en) 2018-01-30
CN102142087A (zh) 2011-08-03

Similar Documents

Publication Publication Date Title
CN104834849B (zh) 基于声纹识别和人脸识别的双因素身份认证方法及系统
CN105468948B (zh) 一种通过社交关系进行身份验证的方法
US10135818B2 (en) User biological feature authentication method and system
US9081947B2 (en) Turing test based user authentication and user presence verification system, device, and method
CN106101136A (zh) 一种生物特征对比的认证方法及系统
WO2018166524A1 (zh) 人脸检测方法和系统、电子设备、程序和介质
CN103607536B (zh) 相机的控制方法及相机
CN105577664B (zh) 密码重置方法及系统、客户端及服务器
CN108959933A (zh) 用于基于风险的认证的风险分析装置及方法
TWI616821B (zh) Bar code generation method, bar code based authentication method and related terminal
JP2013097760A (ja) 認証システム、端末装置、認証プログラム、認証方法
CN103324947B (zh) 一种认证方法及一种鉴权方法
RU2018106036A (ru) Система и способ для подтверждения авторства сеанса электронной подписи
JP2016009453A (ja) 顔認証装置および顔認証方法
EP3807794B1 (en) Age verification
CN108734067A (zh) 一种人证比对的认证方法、系统及相机
WO2014178853A1 (en) Ad-hoc, face-recognition-driven content sharing
US20190073461A1 (en) Methods and systems for ensuring that an individual is authorized to conduct an activity
WO2011095071A1 (zh) 一种用户认证的方法及装置
CN103279764A (zh) 基于人脸识别的网络实名认证系统
US20220046012A1 (en) Method and System for Verifying the Identity of a User
CN107786487B (zh) 一种信息认证处理方法、系统以及相关设备
WO2017041494A1 (zh) 一种信息处理方法及终端、计算机存储介质
CN107437016B (zh) 应用控制方法及相关产品
CN102833182A (zh) 即时通信中进行人脸识别的方法、客户端及系统

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: 11739350

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 1201003926

Country of ref document: TH

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 14/12/2012)

122 Ep: pct application non-entry in european phase

Ref document number: 11739350

Country of ref document: EP

Kind code of ref document: A1