| 摘 要: 为了提高人脸识别的效率,本文提出了一种将小波分析、深度学习和adaboost分类器相结合的人脸识别方 法。传统的基于小波变换的人脸识别算法仅仅提取了小波分解的低频分量用于分类图像的特征,为了更有效地提取人脸 图像特征,提出了一种将传统特征和深度特征相融合的人脸识别算法。首先,通过二维离散小波变换函数对人脸图像进行 二维离散小波变换,提取出人脸图像的低频部分作为特征值,接着通过深度残差网络提取人脸深度特征,最后将融合后的 特征应用adaboost分类器进行分类识别。通过在ORL人脸库实验证明,融合后的方法能有效地提高分类识别率。 |
| 关键词: 小波变换；人脸识别；残差网络；Adaboost分类 |
|
中图分类号: TP393
文献标识码: A
|
|
| AdaBoost Face Recognition Model Based on Deep Learning and Two-dimensional Discrete Wavelet Decomposition |
|
HUANG Jian
|
( School of Electronic and Information Engineering, Anhui Jianzhu University, Hefei 230601, China)
|
| Abstract: In order to improve the efficiency of face recognition,this paper proposes a face recognition method combining wavelet analysis,deep learning and adaboost classifier.The traditional face recognition algorithm based on wavelet transform only extracts the low-frequency components of wavelet decomposition to classify the features of images.In order to extract the features of face images more effectively,a face recognition algorithm combining traditional features and depth features is proposed.Firstly,two-dimensional discrete wavelet transform is applied to face images to extract low-frequency parts of face images as feature values.Then,deep residual network is used to extract face depth features.Finally,the fused features are classified and recognized by adaboost classifier.Through the ORL face database experiment,the fusion method can effectively improve the classification recognition rate. |
| Keywords: wavelet transform;face recognition;residual network;Adaboost classification |
