Title :
Frequency domain feature-based face recognition technique for different poses and low-resolution conditions
Author :
Shahdi, Seyed Omid ; Abu-Bakar, S.A.R.
Author_Institution :
Fac. of Electr. Eng., Univ. Teknol. Malaysia, Skudai, Malaysia
Abstract :
Pose variations are known to give real challenges in face recognition system. In this paper we proposed a method to recognize non-frontal faces with high performance by relying only on single full frontal gallery faces. By utilizing only small regions of the face or patches, we compute the Fourier coefficients of these patches for each image and transform them into a single vector. Hence, instead of comparing and matching pixels values we use these vectors to form a linear relationship which is then used to estimate the frontal face vector and then compare it with the actual frontal feature vector. The results show an average performance accuracy of 90% across all pose.
Keywords :
Fourier transforms; face recognition; frequency-domain analysis; pose estimation; Fourier coefficients; frequency domain feature; frontal face vector estimation; low-resolution conditions; nonfrontal face recognition technique; pose variations; Databases; Face; Face recognition; Feature extraction; Training; Vectors; face recognition; frequency domain; local regions (patches); mapping coefficient; varying pose;
Conference_Titel :
Imaging Systems and Techniques (IST), 2011 IEEE International Conference on
Conference_Location :
Penang
Print_ISBN :
978-1-61284-894-5
DOI :
10.1109/IST.2011.5962222