摘要
提出了一种基于谱直方图和支持向量机的人脸检测算法。首先使用梯度滤波器、LoG滤波器和LBP算子计算图像的谱直方图,然后使用支持向量机进行分类。经实验表明,算法可以克服光照、姿势、表情、眼镜等干扰,并允许有局部遮挡,具有较强的鲁棒性。
A face detection algorithm based on spectral histograms and support vector machines is presented.The spectral histograms of images computed by gradient filters,LoG filters and LBP operators,and classified by support vector machines.Experiment results proved that our algorithm can overcome the variability of lighting condition,pose,facial expression,glasses and occlusion,it is a robust face detection algorithm.
出处
《微计算机信息》
2010年第1期206-208,共3页
Control & Automation
基金
北方民族大学校级科研项目(2006Y025)
关键词
人脸检测
谱直方图
支持向量机
Face detection
spectral histogram
support vector machine
