摘要
稀疏表示分类算法在人脸识别中表现出较好性能,但其对面部遮挡的鲁棒性较差,在特征提取时未利用面部鲁棒特征且实时性差。针对以上问题,本文提出了融合遮挡检测与HOG特征协作表示的人脸识别方法。首先,设计了一种基于HOG特征和支持向量机的遮挡检测方法,根据常见遮挡的分布将人脸分块,在子块上检测遮挡。然后,结合遮挡检测结果,仅提取非遮挡部分的HOG特征并利用协作表示方法进行分类。测试结果表明:对AR人脸库中遮挡图像的平均识别率为95.2%,对AR、Yale人脸库中发生光照和表情变化人脸的识别率分别为97.3%和98.6%,且运算速度比稀疏表示方法提升了约293倍。与基于稀疏表示的分类方法相比,该方法的鲁棒性和实时性都有明显提升。
Sparse representation based classification（SRC）shows good performance in face recognition experiments,but occlusion is still a tough problem because the robust features of face have not been fully utilized.In addition,the real-time performance of SRC is poor.In this paper,a face recognition method with fusion of occlusion detection and collaborative representation based on HOG feature was proposed.Firstly,an occlusion detector based on HOG feature and support vector machine was designed,which could detect the occlusion on the local patches of the face images that were divided according to common occlusion distribution.Then,based on the results of occlusion detection,the HOG feature of the non-occluded face regions was extracted and collaborative representation was adopted for classification.Finally,experiments were conducted on AR and Yale face databases.The average recognition rate of face images with occlusion in AR face database reached 95.2%.The recognition rates of face images with illumination and expression changes in AR and Yale face databases reached 97.3% and 98.6%respectively.The calculation speed of the proposed method was about 293 times faster than SRC.Experimental results demonstrate that the proposed method yields significant performance improvements compared to SRC method.
作者
高秀洋
房胜
李哲
GAO Xiuyang FANG Sheng LI Zhe(College of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao,Shandong 266590,Chin)
出处
《山东科技大学学报（自然科学版）》
CAS
2017年第1期111-117,共7页
Journal of Shandong University of Science and Technology(Natural Science)
基金
国家自然科学基金项目(61502278
61170253)
山东科技大学人才引进科研启动基金项目(2015RCJJ067)
关键词
HOG特征
支持向量机
遮挡检测
协作表示
人脸识别
Histogram of Oriented Gradient（HOG）
support vector machine
occlusion detection
collaborative representation
face recognition
