支持向量机在面部神经麻痹检测中的应用

Examination of Facial Neural Anaesthetization Using Support Vector Machine

通过提取标准人脸正面图像的鼻尖、眼角、鼻角、嘴角等特征点,构成11维特征向量,在介绍支持向量机(SVM)基本原理和实现算法的基础上,通过SVM对大样本的11维特征向量进行学习,将面部神经麻痹的图像从正常人脸图像中分离出来,为医生诊断提供依据。试验证明,该方法对检测面部神经麻痹的准确率是令人满意的。

dimension eigenvectors which are constituted of some feature points such as nose,eyes,mouth and their edges,etc are trained by support vector machine whose principle and arithmetic are introduced.After training,the pictures of neural anaesthetization are separated from the normal ones,which can provide the evidence to doctors.Experimentations have proved that this method was satisfying of the rate of preciseness.