融合HOG与PCA算法的人脸识别

Face Recognition Based on HOG Feature and PCA Algorithm

为了进一步研究人脸识别问题文章融合HOG特征与PCA算法对人脸进行识别研究。首先计算人脸图像的方向梯度直方图(HOG),将输出的每一个特征向量纵向堆叠为一个二维矩阵。然后使用主成分分析(PCA)进行特征降维,减少特征间的相关性和噪声。最后使用支持向量机(SVM)进行分类识别。整个算法模型在ORL人脸数据库中进行实验,最终结果显示识别准确率为96.0%;使用ROC曲线评价该方法的优劣得到曲线下的面积为0.9898。

In order to solve the problem of face recognition,this paper proposes a face recognition method that combines HOG feature and PCA algorithm.Firstly,the directional gradient histogram(HOG)of face images in the database is calculated,and each output eigenvector is stacked into a two-dimensional matrix.Then,principal component analysis(PCA)is used to reduce the dimension of features and reduce the correlation and noise between features.Finally,support vector machine(SVM)is used for classification and recognition.The whole algorithm model is tested in ORL face database,and the final result shows that the recognition accuracy is 96.0%.The ROC curve is used to evaluate the advantages and disadvantages of this method.The area under the curve is 0.9898.