一种基于n元组分类器的人脸图象识别方法

摘要

Human face recognition is a hot point in pattern recognition and AI field, in this paper, a new approach based on continuous n-tuple classifier for multi-grey face image recognition is presented. Sparsely storing technique is used in this algorithm. Experiments show that this algorithm is better than tranditional n-tuple classifier and eigenface in correct proportion of recognition and time of recognition.%人脸图象识别是模式识别和人工智能领域的一个热点，文章介绍了一种可以用于多灰度级人脸图象识别的连续n元组分类器方法，采用稀疏存储等技术，实验结果表明它比传统的特征脸方法和传统的n元组分类器方法简捷、快速，识别正确率高。