文章基本信息

标题：A Novel Feature Descriptor for Face Anti-Spoofing Using Texture Based Method
本地全文：下载
作者：Raghavendra R. J. ; R. Sanjeev Kunte
期刊名称：Cybernetics and Information Technologies
印刷版ISSN：1311-9702
电子版ISSN：1314-4081
出版年度：2020
卷号：20
期号：3
页码：159-176
DOI：10.2478/cait-2020-0035
语种：English
出版社：Bulgarian Academy of Science
摘要：In this paper we propose a novel approach for face anti-spoofing calledExtended Division Directional Ternary Co-relation Pattern (EDDTCP). TheEDDTCP encodes co-relation of ternary edges based on the centre pixel gray valueswith its immediate directional neighbour and its next immediate average directionalneighbour, which is calculated by using the average of cornered neighbours withdirectional neighbours. The proposed method is robust against presentation attacksby extracting the spatial information in all directions. Three Experiments wereperformed by using all the four texture descriptors (LBP, LTP, LGS and EDDTCP)and the results are compared. The proposed face anti-spoofing method performsbetter than LBP, LTP and LGS.
关键词：Face Anti-spoofing; Extended Division Directional Ternary Co-relation Pattern (EDDCP); Texture analysis; Replay-Attack