Learning center-epsilon local binary pattern for bovine IRIS recognition

作     者：Song, Yang Sun, Shengnan Zhao, Lindu 

作者机构：Institute of Systems Engineering Southeast University No. 2 Sipailou Nanjing 210096 China 

出 版 物：《ICIC Express Letters》 (ICIC Express Lett.)

年 卷 期：2013年第7卷第4期

页      面：1209-1214页

核心收录：

主　　题：Local binary pattern 

摘      要：The local binary pattern (LBP) descriptor has gained a lot of interest in recent years due to its simplicity and excellent performance in various applications. Nevertheless, the conventional LBP operator misses the information of the central pixel. In this paper, we propose a novel bovine iris recognition approach using center-epsilon local binary pattern (CΕ-LBP) descriptor. This novel local feature descriptor incorporates the information of the central pixel into the conventional LBP descriptor. Furthermore, the proposed descriptor can tolerate flat image areas, illumination changes and noise, while a variable Ε is introduced to the sign function. The experimental results on the bovine iris database verify the effectiveness of our proposed local feature descriptor. © 2013 ICIC International.