Face Recognition Via Sparse Representation and 2D Feature Extraction

• 作 者: LiNai-qing;WuCheng-zhi;ZhuZhang-qing;ChenChun-lin;
• 机 构: Department of Control and Systems Engineering, School of Management and Engineering, Nanjing Univers
• 出 处: 第17届中国系统仿真技术及其应用学术年会(17th CCSSTA 2016)
• 发表日期: 2016年期

　　A general classification algorithm using compressed sensing has been proposed for pattern recognition.This classifier is also known as sparse representation classification (SRC).It has been proved that SRC could be combined with feature extraction schemes in order to reduce data dimensions and computational complexity.Among many feature extraction schemes, 2D subset learning techniques (2DSTs) have several advantages over 1D subset learning techniques (1DSTs).The 2DSTs not only have much lower time complexity but also can preserve the structures of images.Hence in this paper, we combine SRC with 2DSTs for further reduction of computational complexity for face recognition.