Texture analysis using local ternary pattern for face anti-spoofing
This paper proposes a new face anti-spoofing approach based on analysis of texture characteristics. Photo images are used for spoofing the face recognition and verification system. These photos are similar to the images of a live person which are exhibiting quite different contrast and textur...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Publications International
2016
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/52944/ http://psasir.upm.edu.my/id/eprint/52944/1/Texture%20analysis%20using%20local%20ternary%20pattern%20for%20face%20anti-spoofing.pdf |
| _version_ | 1848852179980386304 |
|---|---|
| author | Parveen, Sajida Syed Ahmed, Sharifah Mumtazah Hasan Abbas, Nidaa Naeem, Nadeem Hanafi, Marsyita |
| author_facet | Parveen, Sajida Syed Ahmed, Sharifah Mumtazah Hasan Abbas, Nidaa Naeem, Nadeem Hanafi, Marsyita |
| author_sort | Parveen, Sajida |
| building | UPM Institutional Repository |
| collection | Online Access |
| description | This paper proposes a new face anti-spoofing approach based on analysis of texture characteristics. Photo images are used for spoofing the face recognition and verification system. These photos are similar to the images of a live
person which are exhibiting quite different contrast and
texture characteristics when compared to real face images. Various feature extraction methods for texture classification including local Binary Patterns emerge as one of the most popular method because of its simplicity and classification accuracy. However, in homogenous regions, the order of the pixel with respect to its neighbors is quite noisy which can affect the performance of Local Binary Pattern. This paper demonstrates the use of local Ternary Pattern (LTP) in face liveness detection system to overcome this problem. The LTP approach is tested on three publicly available NUAA Photograph Imposter database, CASIA Face Anti
-Spoofing Database and REPLAY-ATTACK database. Moreover, different experiments are performed by applying different sizes of neighbor pixels and radius of the patches. The test results are compared with the LBP operator and other state-of art work. The proposed face anti-spoofing method performs better than conventional texture based methods. |
| first_indexed | 2025-11-15T10:33:58Z |
| format | Article |
| id | upm-52944 |
| institution | Universiti Putra Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T10:33:58Z |
| publishDate | 2016 |
| publisher | Publications International |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | upm-529442017-11-22T10:52:04Z http://psasir.upm.edu.my/id/eprint/52944/ Texture analysis using local ternary pattern for face anti-spoofing Parveen, Sajida Syed Ahmed, Sharifah Mumtazah Hasan Abbas, Nidaa Naeem, Nadeem Hanafi, Marsyita This paper proposes a new face anti-spoofing approach based on analysis of texture characteristics. Photo images are used for spoofing the face recognition and verification system. These photos are similar to the images of a live person which are exhibiting quite different contrast and texture characteristics when compared to real face images. Various feature extraction methods for texture classification including local Binary Patterns emerge as one of the most popular method because of its simplicity and classification accuracy. However, in homogenous regions, the order of the pixel with respect to its neighbors is quite noisy which can affect the performance of Local Binary Pattern. This paper demonstrates the use of local Ternary Pattern (LTP) in face liveness detection system to overcome this problem. The LTP approach is tested on three publicly available NUAA Photograph Imposter database, CASIA Face Anti -Spoofing Database and REPLAY-ATTACK database. Moreover, different experiments are performed by applying different sizes of neighbor pixels and radius of the patches. The test results are compared with the LBP operator and other state-of art work. The proposed face anti-spoofing method performs better than conventional texture based methods. Publications International 2016 Article PeerReviewed application/pdf en http://psasir.upm.edu.my/id/eprint/52944/1/Texture%20analysis%20using%20local%20ternary%20pattern%20for%20face%20anti-spoofing.pdf Parveen, Sajida and Syed Ahmed, Sharifah Mumtazah and Hasan Abbas, Nidaa and Naeem, Nadeem and Hanafi, Marsyita (2016) Texture analysis using local ternary pattern for face anti-spoofing. Science International, 28 (2). pp. 965-971. ISSN 1013-5316 http://www.sci-int.com |
| spellingShingle | Parveen, Sajida Syed Ahmed, Sharifah Mumtazah Hasan Abbas, Nidaa Naeem, Nadeem Hanafi, Marsyita Texture analysis using local ternary pattern for face anti-spoofing |
| title | Texture analysis using local ternary pattern for face anti-spoofing |
| title_full | Texture analysis using local ternary pattern for face anti-spoofing |
| title_fullStr | Texture analysis using local ternary pattern for face anti-spoofing |
| title_full_unstemmed | Texture analysis using local ternary pattern for face anti-spoofing |
| title_short | Texture analysis using local ternary pattern for face anti-spoofing |
| title_sort | texture analysis using local ternary pattern for face anti-spoofing |
| url | http://psasir.upm.edu.my/id/eprint/52944/ http://psasir.upm.edu.my/id/eprint/52944/ http://psasir.upm.edu.my/id/eprint/52944/1/Texture%20analysis%20using%20local%20ternary%20pattern%20for%20face%20anti-spoofing.pdf |