基于混合核函数支持向量机和遗传算法的人脸识别

摘要

提出了一种基于混合核函数支持向量机和遗传算法的识别方法,用于人脸识别.该方法结合了支持向量机的学习性能和遗传算法的寻优性能,与传统的方法相比,具有速度快、误差少、效率高的特点,在实验中能够较精确地对人脸进行识别.%This paper proposes a recognition method, which is based on mixed kernel function support vector machine and genetic algorithm, for human face recognition. The method combines the learning performance of support vector machine with the searching performance of genetic algorithm. Compared with traditional methods, this face recognition method has the features of faster, less error and more efficient, in experiment the meth is more accurate in recognising human faces.