一种新的Gabor小波支持向量机人脸识别算法

Novel Gabor wavelet support vector machines algorithm for human face recognition

提出了一种新的基于Gabor小波特征重组的支持向量机人脸识别方法。该方法首先计算5个尺度和8个方向的Gabor小波变换结果,再把不同人脸中的同一尺度和方向的变换结果进行特征重组,得到40个新特征矩阵,分别利用PCA方法降维去噪,最后构造40个支持向量机分类器并采用选票决策机制决定识别结果。实验结果表明,该方法不仅拓宽了主元分析法中累积方差贡献率可选范围,并在一定程度上解决了核参数选择难的问题,同时取得了理想识别效果。

A novel support vector machines algorithm for human face recognition based on Gabor wavelet features reorganization was proposed. Firstly, the Gabor wavelets transformation results including five scales and eight directions were calculated and 40 feature matrices which were reconstrueted with the same scale and the same direction transform results of the different face images were obtained. Secondly, the dimensionality reduction and denoised technique with PCA was applied to form the new training samples. Lastly, 40 SVMs classifiers were constructed and the vote decision strategy was used to determine the recognition result. The experimental results show that the proposed method expands the selectable range of the variance contribution rate in PCA method and the difficult problem was settled to select the kernel parameters in the certain stage. At the same time, the ideal recognition rate was obtained.