摘要
为了进一步提高支持向量机分类器人脸识别率和识别速度,提出了采用分块局部二值模式(B-LBP)特征和麻雀搜索算法(SSA)优化支持向量机(SVM)的人脸识别算法。算法首先对人脸图像分块并提取每个子分块统一化局部二值模式特征,通过特征融合、快速PCA降维处理后得到人脸图像特征;再使用SSA优化SVM分类器的核函数参数和惩罚系数,提高分类器的分类精度和速度;最后将人脸图像特征输入到优化后的SVM分类器进行分类。实验结果表明,与其他几种人脸识别算法相比,基于B-LBP特征和SSA优化SVM的人脸识别算法的人脸识别率和识别速度显著提升。
In order to further improve the face recognition rate and recognition speed of the support vector machine classifier,a face recognition algorithm based on optimized Support Vector Machine(SVM)by Block Local Binary Pattern(B-LBP)and Sparrow Search Algorithm(SSA)is proposed.The algorithm first partitions the face image and extracts the unified local binary pattern features of each sub-partition,and obtains the face image features through feature fusion and fast PCA dimensionality reduction processing;and then uses SSA to optimize the kernel function parameters and penalty coefficients of the SVM classifier,to improve the classification accuracy and speed of the classifier;finally,the face image features are input into the optimized SVM classifier for classification.The experimental results show that the face recognition rate and recognition speed of the algorithm are significantly improved.
作者
朱强军
许佳炜
王杨
张广海
高丽
ZHU Qiangjun;XU Jiawei;WANG Yang;ZHANG Guanghai;GAO Li(Department of Electronic Engineering,Wanjiang College of Anhui Normal University,Wuhu Anhui 241008;School of Computer and Information,Anhui Normal University,Wuhu Anhui 241003)
出处
《湖北理工学院学报》
2022年第6期1-5,38,共6页
Journal of Hubei Polytechnic University
基金
安徽省高校优秀拔尖人才培育项目(项目编号:gxyq2020093)
安徽师范大学皖江学院教学质量工程项目(项目编号:WJXGK-202201)。
关键词
人脸识别
分块局部二值特征
支持向量机
麻雀搜索算法
特征融合
face recognition
Block Local Binary Pattern
Support Vector Machine
Sparrow Search Algorithm
feature fusion
