摘要
以二维Gabor小波变换提取人脸图像特征作为全局特征,对图像进行不等分块,人脸图像区域所在块加大权重,并提取每个子块的特征作为局部特征,对全局特征结合局部特征采用DCT进行降维处理,用支持向量机分类模型进行人脸识别。实验表明:与较同等分块的识别算法相比,该算法可提高人脸识别率。
Gabor wavelet transform two-dimensional image feature extraction lace as a global feature,the image is unequal block, where the blocks face image area to increase the weight, and extract the characteristics ol each sub-block as the local characteristics,global characteristics ol each sub-block as the local characteristics, global characteristics combined with local characteristics use D C T dimension reduction,face recognition algorithm can improve the recognition rate.
作者
刁琦
张江楠
古丽米拉.克孜尔别克
DIAO Qi;ZHANG Jiang-nan;Gulimila .kezierbieke(School of Computer and Information Engineering,Xinjiang Agricultural University,Urumqi 830052,China)
出处
《信息技术》
2016年第12期36-38,43,共4页
Information Technology
基金
新疆农业大学大学生创新项目(201510758106)
