基于稀疏表示的人脸识别算法研究

Research on Face Recognition Algorithm Based on Sparse Representation

本文使用了全新的图像表示方法用于生成虚拟图像,并根据稀疏表示的思想实现了一个图像分类算法。首先,在分类测试样本时,使用图像分类算法分别获得原始图像和虚拟图像对应的分类得分;其次,使用一种简单高效的得分融合方案获得测试样本的最终分类得分;最后,实现测试样本的分类。本文将该图像分类算法应用于人脸识别,并在多个公开的人脸数据库上进行了实验。实验结果表明,本文提出的算法在人脸识别准确率方面取得了非常有竞争力的结果。

In this paper,a new image representation method is used to generate virtual images,and an image classification algorithm is implemented according to the idea of sparse representation.Firstly,when classifying the test samples,the image classification algorithm is used to obtain the classification scores corresponding to the original image and the virtual image respectively;Secondly,a simple and efficient score fusion scheme is used to obtain the final classification score of the test samples;Finally,the classification of test samples is realized.In this paper,the image classification algorithm is applied to face recognition,and experiments are carried out on several public face databases.Experimental results show that the proposed algorithm has achieved very competitive results in face recognition accuracy..