Local binary pattern (LBP) and local phase quantization (LBQ)
based on Gabor filter for face representation
Shu-Ren Zhou
a,b,
n
, Jian-Ping Yin
a
, Jian-Ming Zhang
b
a
Postdoctoral Station of Computer School, National University of Defense Technology, Changsha, Hunan 410073, China
b
Computer & Communication Engineering School, Changsha University of Science & Technology, Changsha 410004, China
article info
Available online 2 November 2012
Keywords:
Face representation
Gabor filter
Local binary pattern
Local phase quantization
Feature fusion
abstract
Sometimes realistic face representation is confronted with blur or low-resolution face images, as a result,
existing classification methods are not powerful and robust enough. This paper proposes a novel face
representation appr oach (GLL) which fuses Gabor filter, Local Binary Pattern (LBP) and Local Phase
Quantization (LPQ). In the process of Gabor filter, it uses Gabor wavelet functions with two scales and eight
orientations to capture the salient visual properties in face image. On this basis of Gabor features, we acquire
LBP features and LPQ features, respectively, so as to fully explore the blur invariant property and the
information in the spat ial domain and among different scales and orientations. Experiments on both CMU-PIE
and Yale B demonstrate the effectiveness of our GLL when dealing with different condition face data sets.
& 2012 Elsevier B.V. All rights reserved.
1. Introduction
Face recognition plays an important role in many applications,
such as human computer interface, visual surveillance, access
control, law enforcement, and more generally image representa-
tion [1]. With the complexity and nonrigidity of face, face
representation has become a verification benchmark for various
pattern recognition algorithms [2].
Various algorithms have been developed and tested for face
representation. Among them, dynamic texture-based approach
[3] is proposed to the recognition of facial action units. The
recently proposed local binary pattern (LBP) operator [4] has
been successfully applied to facial expression [5] and face
recognition [6]. Local phase quantization (LPQ) is a novel method
which is used for recognition of blurred faces [7], LPQ is based on
quantizing the Fourier transform phase in local neighborhoods, in
face image analysis, histograms of LPQ labels computed within
local regions are used as a face descriptor similarly to the widely
used LBP methodology for face image description. In [8], Marsico
et al. provide FAce Recognition against Occlusions and Expression
Variations (FARO) as a new method based on partitioned iterated
function systems (PIFSs), which is quite robust with respect to
expression changes and partial occlusions. Recently, the fusion
method combined Gabor filter has been key focus to many
researchers [9–11]. Gabor wavelets capture the local structure
corresponding to specific spatial frequency, spatial locality, and
selective orientation which are demonstrated to be discriminative
and robust to illumination and expression changes. Lei et al.
provide a face representation and recognition approach by com-
bined the Gabor filters with multiscale and multiorientation to
local binary pattern analysis [12].
In this paper, we propose a novel face represe ntation method
which explores not only the blur invariant property, b ut also the
information in the spatial domain and among different scales and
orientations. First, the multiscale and multiorientation repr esenta-
tions are derived by convolving the face image with a Gabor filter
bank and formulated as serial transformed images. Second, the LPQ
labels [13] computed within local regions are used as a face descriptor
for face transformed image description. Third, LBP operator is applied
on the face transformed image of Gabor filter. In this way, we encode
the neighboring information and the texture information not only in
image space but also among different scales and orientations of Gabor
faces. Finally, the final feature vectors are acquired from the fusion
transformed images by concatenated histograms.
We introduce the basic approach of face representation in
Section 2, containing Gabor filter, LBP and LPQ. Then Section 3
describes the experimental results and Section 4 finally presents
the concluding remarks.
2. Face representation
2.1. Gabor wavelets
Gabor kernels are similar to the receptive field profiles in
cortical simple cells, which are characterized as localized,
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/neucom
Neurocomputing
0925-2312/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.neucom.2012.05.036
n
Corresponding author at: Computer & Communication Engineering School,
Changsha University of Science & Technology, Changsha 410004, China.
Tel.: þ86 731 8525 8462.
E-mail address: zsr@csust.edu.cn (S.-R. Zhou).
Neurocomputing 116 (2013) 260–264