基于双重特征空间的人脸识别

Face recognition based on double characteristic space

多目标遗传算法(MOGA)是求解多目标优化问题的有效工具,因而在求解实际问题中得到越来越广泛的应用。PCA是一种基于二阶统计的最小均方误差意义上的最优维数压缩技术,PCA方法所抽取特征的各分量之间是统计不相关的。在人脸识别的实际应用中,将多目标遗传算法引入到PCA所生成的特征空间的优化中,提出基于双重特征空间的人脸识别算法。通过对剑桥ORL库实验表明,该方法与传统的PCA相比,识别率得到明显提高。

Multi-objective genetic algorithm （MOGA） is an effective way to solve multi-objective optimization problem and is used more and more extensively in solution of practical problem. PCA （principal component analysis） is the optimal dimension compression technique based on second-order information, in the sense of mean-square error. Features extracted by PCA are statistically uncorrelated to each other. In the applications on face image recognition, optimization of characteristic space which is produced by PCA is introduced. A new face recognition method based on double characteristic space is proposed. The experiment to ORL face library shows that the new method provide satisfactory recognition performance.