摘要
文章提出了一种基于聚合图像的人脸识别方法,首先提取人脸数据库中多张人脸图像的变化特征,叠加至训练样本集上的每个人脸样本图像上,形成聚合图像。而后,利用FAST检测器检测图像的特征点,采用具有加速鲁棒性特征的SURF算法提取和描述特征点,得到人脸图像的特征向量集,并引入AMM主动表观模改进SURF匹配算法,从而基于训练与测试图像的特征点匹配,实现人脸识别。最后经实验对比分析,该方法在提取识别效用较大的特征点、识别率上均存在应用优势,且使用聚合图像可有效提升人脸的识别率。
A face recognition method based on aggregated images is proposed. Firstly, the features of multiple face images in the face database were extracted and superimposed on each face sample image in the training sample set to form aggregated images. Then, the FAST detector was used to detect the feature points of the image, the SURF algorithm with accelerated robustness was used to extract and describe the feature points, and the feature vector set of the face image was obtained. The AMM active appearances model was introduced to improve the SURF matching algorithm, so that the face recognition could be realized based on the matching of the feature points of the training and testing images. Finally, through the comparative analysis of experiments, this method has advantages in extracting feature points with high recognition efficiency and recognition rate, and the use of aggregated images can effectively improve face recognition rate.
作者
洪雨天
杨梓超
Hong Yutian;Yang Zichao(Guangdong Power Grid Co., Ltd. Guangzhou, 510000, China)
出处
《现代科学仪器》
2019年第1期41-44,共4页
Modern Scientific Instruments
关键词
聚合图像
特征描述算子
改进SURF
人脸识别
Aggregated image
feature description operator
improved SURF
face recognition
