首页> 外国专利> IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, PERSON IDENTIFICATION APPARATUS, AND METHOD AND PROGRAM OF PRODUCING/UPDATING DICTIONARY DATA IN PERSON IDENTIFICATION APPARATUS

IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, PERSON IDENTIFICATION APPARATUS, AND METHOD AND PROGRAM OF PRODUCING/UPDATING DICTIONARY DATA IN PERSON IDENTIFICATION APPARATUS

机译:图像处理装置,图像处理方法,人员识别装置以及在人员识别装置中产生/更新字典数据的方法和程序

摘要

An image processing apparatus, an image processing method, a person identification apparatus, and a dictionary data generating & updating method and program in the person identification apparatus are provided to select information with higher reliability on the basis of the reliability information and to use the selected information in processing. A face recognition dictionary database(10) stores person dictionary data of each of plural persons registered for being identified via face image in association with personal identification information. A face recognition image processor(11) detects a face and eyes from an original image obtained via an image data inputting interface(7) and rotates the face image for putting the eyes along a horizontal line. A face recognition image processor(12) plots feature points for extracting feature data of feature portions like eyebrows, eyes or a nose in the face image, detects a feature vector as feature data of the captured face image, and provides the feature vector for a face recognition processor. The face recognition processor compares the received feature vector with face recognition dictionary data of plural persons stored at the database, and detects face recognition dictionary data identical to the feature vector. A score calculator(14) calculates a score, which indicates the reliability of a face image extracted from the captured image through the image data inputting interface, when receiving data as to the face and the eyes detected from the face recognition image processor.
机译:提供一种图像处理设备,图像处理方法,人员识别设备以及词典数据生成和更新方法以及该人员识别设备中的程序,以基于可靠性信息选择具有较高可靠性的信息并使用所选择的处理中的信息。面部识别字典数据库(10)与个人识别信息相关联地存储登记为经由面部图像被识别的多个人物中的每个人物的人物字典数据。面部识别图像处理器(11)从经由图像数据输入接口(7)获得的原始图像中检测面部和眼睛,并旋转面部图像以使眼睛沿着水平线。面部识别图像处理器(12)绘制特征点以提取面部图像中的眉毛,眼睛或鼻子等特征部分的特征数据,将特征向量检测为所捕获的面部图像的特征数据,并为人脸识别处理器。面部识别处理器将接收到的特征向量与存储在数据库中的多个人的面部识别字典数据进行比较,并检测与特征向量相同的面部识别字典数据。得分计算器(14)计算得分,该得分表示当接收到从面部识别图像处理器检测到的关于面部和眼睛的数据时,通过图像数据输入接口从捕获图像提取的面部图像的可靠性。

著录项

  • 公开/公告号KR20080042009A

    专利类型

  • 公开/公告日2008-05-14

    原文格式PDF

  • 申请/专利权人 SONY CORPORATION;

    申请/专利号KR20070113084

  • 发明设计人 SHEN YUHANG;KOJIMA TAMAKI;EURA YUKA;

    申请日2007-11-07

  • 分类号G06K9/68;G06K9/60;G06T7/40;

  • 国家 KR

  • 入库时间 2022-08-21 19:53:42

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号