公开(公告)号：EP2864933A1

公开(公告)日：2015-04-29

申请号：EP13808762.2

申请日：2013-04-24

申请人： Nokia Corporation

发明人： MA, Yong ,  ZOU, Yan Ming ,  WANG, Kong Qiao

IPC分类号： G06K9/00 ,  G06K9/62

CPC分类号： G06K9/00281 ,  G06K9/00261 ,  G06K9/00288

摘要： The present invention provides a method for obtaining a human-face feature vector from a video image sequence, comprising: detecting a same human-face image in a plurality of image frames of the video sequence; dividing the detected human-face image into a plurality of local patches with a predetermined size, wherein each local patch is around or near a human-face feature point; determining a correspondence relationship between respective local patches of the same human-face image in the plurality of image frames of the video sequence; and using human-face local feature vector components extracted from respective local patches having a mutual correspondence relationship to form human-face local feature vectors representing facial points corresponding to the local patches. Besides, the present invention further provides an apparatus for obtaining a human-face feature vector from a video image sequence and a corresponding computer program product.

摘要翻译： 本发明提供一种从视频图像序列获得人脸特征向量的方法，包括：检测视频序列的多个图像帧中的相同的人脸图像; 将检测到的人脸图像划分成具有预定大小的多个局部斑块，其中每个局部斑块围绕或靠近人脸特征点; 确定所述视频序列的所述多个图像帧中的相同人脸图像的各个局部斑块之间的对应关系; 并且使用从具有相互对应关系的各个局部片段提取的人脸局部特征向量分量来形成表示对应于局部斑块的面部点的人脸局部特征向量。 此外，本发明还提供了一种用于从视频图像序列和相应的计算机程序产品获得人脸特征向量的装置。

公开(公告)号：EP2825996A1

公开(公告)日：2015-01-21

申请号：EP12871162.9

申请日：2012-03-13

申请人： Nokia Corporation

发明人： ZOU, Yanming ,  MA, Yong ,  WANG, Kongqiao

IPC分类号： G06K9/00

CPC分类号： G06K9/00295 ,  G06K9/00281 ,  G06K9/00288 ,  G06K9/00302 ,  G06K9/4642

摘要： A method, apparatus and computer program product for an improved facial recognition system are provided. Some embodiments may utilize a weighted block division of an image and capture a property measurement for pixels residing within a block. The measurements may be converted to vectors, compressed, and compared against compressed vectors of enrolled images to identify a characteristic or an image of a matching subject. Training processes may be utilized in order to optimize block divisions and weights.