中文摘要：

基于L1范数的特征提取算法因其识别率高的优点,使图像识别能力有了显著的提升,在图像识别领域有着重要的应用.本文对L1范数特征提取方法的发展历程进行了分析,讨论了基于L1范数PCA、LDA及它们对应的各种改进型的研究现状,通过人脸数据库实验测试了相关L1范数算法的识别性能优劣.最后对各种方法的应用前景做了展望.

英文摘要：

The feature extraction algorithm based on L1 norm,which has received extensive attention in recent years,has significantly promoted image recognition ability for its high recognition rate,and it has important application in image recognition.This paper analyzes the development history of L1 norm feature extraction algorithm and discusses the present research status of L1-norm based on PCA and LDA and their corresponding various improved versions.Their recognition performance is demonstrated by experiments on face image databases.Finally this paper makes expectation on research tendency in this field.