Recent Advancements in Gabor Wavelet-Based Face Recognition

Recent Advancements in Gabor Wavelet-Based Face Recognition

Iqbal Nouyed, M. Ashraful Amin
Copyright: © 2015 |Pages: 13
DOI: 10.4018/978-1-4666-5888-2.ch025
(Individual Chapters)
No Current Special Offers

Chapter Preview



Due to their biological relevance (Daugman, 1980; Marcelja, 1980) and computational properties Gabor wavelets were introduced to image analysis. As a feature generator Gabor filters are widely used in face recognition. .

Since the kernels of Gabor wavelets are similar to the 2D receptive field profiles of the mammalian cortical simple cells, they exhibit desirable characteristics of spatial locality and orientation selectivity. Also they are optimally localized in the space and frequency domains. The Gabor wavelets (kernels / filters) can be defined as following, (Lades, et al., 1993)

(1) where978-1-4666-5888-2.ch025.m02 and 978-1-4666-5888-2.ch025.m03 define the scale and orientation of the Gabor kernel, 978-1-4666-5888-2.ch025.m04denotes the pixel, i.e., 978-1-4666-5888-2.ch025.m05;978-1-4666-5888-2.ch025.m06 denotes the Euclidean norm operator, and the wave vector 978-1-4666-5888-2.ch025.m07 is defined as:
(2) where978-1-4666-5888-2.ch025.m09 is the orientation parameter and 978-1-4666-5888-2.ch025.m10, here978-1-4666-5888-2.ch025.m11is the spacing factor between filters in the frequency domain.

Usually, Gabor filters at five scales (978-1-4666-5888-2.ch025.m12in Equation (1), and eight orientations (978-1-4666-5888-2.ch025.m13ranging between 978-1-4666-5888-2.ch025.m14to 978-1-4666-5888-2.ch025.m15in Equation (2) are applied on each preprocessed facial image. These 40 filters are shown in Figure 1.

Figure 1.

Real part of Gabor kernels: 8 orientations (978-1-4666-5888-2.ch025.m16) and 5 scales (978-1-4666-5888-2.ch025.m17)


Key Terms in this Chapter

Primary Visual Cortex: The part of the cerebral cortex responsible for processing visual information.

Face Authentication: Concerned with validating a claimed identity based on the image of a face.

Principal Component Analysis (PCA): An orthogonal linear transformation that transforms the data to a new coordinate system preferably lower dimensional than original.

Local Gabor Binary Pattern Histogram Sequence (LGBPHS): A non-statistics based face representation approach in which training procedure is unnecessary to construct the face model.

Support Vector Machine (SVM): A supervised learning model with associated learning algorithms that analyze data and recognize patterns.

Gabor Wavelet: A set of linear filters used to extract features from multidimensional signal.

Face Recognition: Biometric identification by scanning a person's face and matching it against a library of known faces.

Complete Chapter List

Search this Book: