公开(公告)号：US20140072221A1

公开(公告)日：2014-03-13

申请号：US14024024

申请日：2013-09-11

Applicant: OMRON Corporation

Inventor： Shun Sakai ,  Motoo YAMAMOTO ,  Hiroyuki Tanaka ,  Tatsuya Murakami ,  Yuki HANZAWA ,  Takahiro Takayama

IPC: G06K9/32 ,  G06T7/00

CPC classification number: G06K9/3233 ,  G06T7/0002 ,  G06T7/11 ,  G06T7/194 ,  G06T2207/10004 ,  G06T2207/30168 ,  G06T2207/30201 ,  H04N5/23222 ,  H04N5/23229

Abstract: An image processing device selects an image in which sharpness of a photographic subject is relatively high compared with sharpness of a background. The image processing device includes a photographic subject detection unit that identifies a photographic subject region on an image, which includes a predetermined photographic subject, by detecting the predetermined photographic subject from the image, a background region identifying unit that identifies a background region on the image, which is other than the photographic subject region, a sharpness identifying unit that identifies sharpness of the photographic subject region and sharpness of the background region, a score calculation unit that calculates a score based on a difference in the sharpness between the photographic subject region and the background region, and a determination unit which determines the image as an optimum image when the score indicates that photographic subject region is clearer than the background region.

Abstract translation: 图像处理装置选择摄影对象的清晰度与背景的锐度相比较高的图像。 图像处理装置包括：摄影对象检测单元，通过从图像中检测预定的拍摄对象来识别包括预定拍摄对象的图像上的被摄体区域，识别图像上的背景区域的背景区域识别单元 不同于摄影对象区域的清晰度识别单元，其识别摄影对象区域的锐度和背景区域的锐度;乐谱计算单元，其基于摄影对象区域和摄影对象区域之间的锐度的差异来计算得分， 背景区域，以及确定单元，当所述得分指示所述摄影对象区域比所述背景区域更清晰时，所述确定单元将所述图像确定为最佳图像。

公开(公告)号：US20130243272A1

公开(公告)日：2013-09-19

申请号：US13774200

申请日：2013-02-22

Applicant: OMRON CORPORATION

Inventor： Yuki HANZAWA ,  Atsushi IRIE ,  Motoo YAMAMOTO

IPC: G06F21/32

CPC classification number: G06F21/32 ,  G06K9/00288 ,  G06K9/68

Abstract: An authentication apparatus comprises an extraction part configured to extract a feature of a face from a face image of a user; a plurality of classifiers configured to identify the user by determining whether or not the feature extracted by the extraction part matches a feature of a face stored in a storage apparatus; a deriving part configured to derive a specific condition based on at least one of the feature extracted by the extraction part and information for authenticating the user; a selecting part configured to select at least one of the classifiers based on the specific condition; and an authentication part configured to authenticate the user when the selected classifier determines that the feature extracted by the extraction part matches the feature of the face stored in the storage.

Abstract translation: 认证装置包括：提取部，被配置为从用户的脸部图像提取脸部特征; 多个分类器，其被配置为通过确定由提取部分提取的特征与存储在存储装置中的面部的特征是否匹配来识别用户; 导出部，被配置为基于由所述提取部提取的特征中的至少一个和用于认证所述用户的信息来导出特定条件; 选择部件，被配置为基于所述特定条件选择所述分类器中的至少一个; 以及认证部，被配置为当所选择的分类器确定由提取部提取的特征与存储在存储器中的面部的特征相匹配时对用户进行认证。