图像灰度熵特征与SVM分类结合的人脸检测

提出了一种基于支持向量机并综合图像灰度值和灰度熵的人脸检测方法．在信息论的编码熵的基础上提出了灰度熵的概念，并把人脸图像的行像素灰度熵和列像素灰度熵作为人脸图像的部分特征．在人脸检测系统中，把人脸图像的灰度值和行像素灰度熵和列像素灰度熵作为特征用支持向量机进行训练，得到检测用分类模型，然后把分类模型应用于人脸检测．实验证明了这种方法的有效性。

Face detection with image gray entropy features and SVM classifier

A way of face detection integrating image gray entropy features and SVM classifier is presented.Based on the concept of encoding entropy in information theory,the concept of gray-scale entropy is introduced and the row pixel entropy and column pixel entropy of faces are combined and proposed as face image features.In the proposed face detection system, the gray-scale value,the row pixel entropy and column pixel entropy of faces are trained by SVM as face features to obtain a classifying model for detection.And the classifying model is then used for face detection.Experiment results demonstrate the feasibility of this way.