摘要
针对脸部有遮挡的人脸识别问题,本文提出将包含眉、眼的子区域作为训练图像与测试图像,同时,为减小光照对识别率的影响,将训练或测试图像调整为零均值图像,最后,本文提出了基于眉眼子区域和零均值预处理的特征脸识别算法;实验结果表明,本文算法的识别准确率较高,在训练图像为1张和6张的情况下,识别准确率分别达到60%和72%。
To address the problem of face recognition with occlusion,this paper proposes to take the eyebrow and eye sub region as the training image and test image.Meanwhile,the training or test images are adjusted to zero mean images to reduce the influence of illumination on the recognition.Finally,this paper proposes an eigenface face recognition algorithm based on eyebrow and eye sub region and zero mean preprocessing.The experimental results show that the recognition accuracy of the proposed algorithm is high.The recognition accuracy can achieve 60%and 72%when one training image and six training images are provided,respectively.
作者
李咏豪
LI Yong-hao(College of Computer Science and Engineering,Nanjing University of Science&Technology,Nanjing Jiangsu 210094)
出处
《数字技术与应用》
2020年第12期102-104,共3页
Digital Technology & Application
关键词
脸部遮挡
人脸识别
光照
识别准确率
face occlusion
face recognition
illumination
recognition accuracy
