摘要
人脸基准点定位可应用于人脸识别、疲劳检测等领域。针对人脸基准点定位中常用的主动表观模型(AAM)的局限性,提出了Haar分类器和AAM算法相结合的人脸基准点定位方法。先是计算图像积分图,然后采用基于Haar特征的AdaBoost级联检测器快速定位出人脸区域,最后将检测到的位置和图像信息传递给AAM进行人脸基准点定位。该方法在抽取的AFLW(annotated facial landmarks in the wild)人脸测试集上表现出良好的性能。实验结果表明,采用该方法能准确、快速定位出人脸基准点。
Facial mark localization is used in face recognition,fatigue detection and other fields.To overcome the limitation of the active appearance model(AAM) which is commonly used in the positioning of face reference points,a facial landmark localization method based on Haar classifier and AAM algorithm is proposed in this paper.Firstly,the integral image of the face image is calculated,and then the AdaBoost cascade detector based on Haar features quickly locates the face region.Finally,the detected face position and image information are passed to the AAM for the positioning of the face reference point.This method shows a good performance on the extracted AFLW(annotated facial landmarks in the wild) face test set.Experimental results show that this method can accurately and quickly locate the face reference point.
作者
程培培
陈典典
马军山
CHENG Peipei;CHEN Diandian;MA Junshan(School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China)
出处
《光学仪器》
2018年第6期48-53,共6页
Optical Instruments
